SOIL 'IICS

Basic Concepts and

EngineeringApplications

II ]1
II

r-

\342\200\242 \342\200\242 ,,,)))

 \342\200\242

Soil Mechamcs:

Basic Concepts and

Engineering Applications

A. AYSEN

The University of Southern Queensland Australia

A.A. BALKEMA PUBLISHERS / LISSE / AB1NGDON/ EXTON(PA) / TOKYO)))

Library of Congress Cataloging-in-PublicationData

Applied for: 5 April 2002

Cover design: Studio Jan de Boer,Amsterdam, The Netherlands.

Printed by: Gorter, Steenwijk,The Netherlands.

Copyright 2002
\302\251 Swets & Zeitlinger B.V., Lisse, The Netherlands

All rights reserved. No part of this publication or the information contained herein
may be reproduced, stored in a retrieval system, or transmitted in anyform or by

any means, electronic, mechanical, by photocopying, recording or otherwise,

without written prior permission from the publishers.

Although all care is taken to ensure the integrity and quality of this publication and
the information herein, no responsibility is assumed by the publishers nor the author
for any damage to property or persons as a result of operation or use of this
publication andCr the information contained herein.

Published by: A.A. Balkema Publishers,a member of Swets & Zeitlinger Publishers
www.balkema.nland www.szp.swets.nl

ISBN 90 5809 358 l)))

Contents

Preface ix

CHAPTER 1 Nature of Soils, Plasticity and Compaction 1

1.1 INTRODUCTION l
1.2 NATURE AND CHEMISTRY OF SOILS l
1.3 MASS-VOLUME RELATIONSHIPS 12
1.4 PARTICLE SIZE DISTRIBUTION 20
1.5 INDEX PROPERTIES AND VOLUME CHANGE IN FINE GRAINED SOILS 26
1.6 SOILCLASSIFICATION FOR GEOTECHNICAL PURPOSES 33
1.7 COMPACTION 41
\177.8 PROBLEMS \177l

1.9 REFERENCES 53

CHAPTER 2 Effective Stress and Pore Pressure in Saturated Soils 55

2,i INTRODUCTION 55
2.2 STATE OF STRESS AT A POINT DUE TO SELF-WEIGHT 55
2.3 STATEOF STRESS AT A POINT DUE TO EXTERNAL FORCES 64
2,4 PROBLEMS 67
2.5 REFERENCES 68

CHAPTER 3 The Movement of Water through Soil 69

3.1 INTRODUCTION 69
3.2 PRINCIPLES OF FLOW IN POROUS MEDIA 69
3.3 PERMEABILITY 72
3.4 FLOW NETS 85
3.5 MATHEMATICS OF THEFLOW IN SOIL 92
3,6 SEEPAGE THROUGHEARTH DAMS 100
3,7 PROBLEMS 107
3.8 REFERENCES 110)))
vi Contents

CHAPTER 4 Shear Strength of SoilsandFailure Criteria 111

4.1 INTRODUCTION \177II

4.2 MOHR-COULOMB FAILURE CRITERION 111

4.3 LABORATORY SHEAR STRENGTH TESTS 118
4.4 STRESS-STRAIN BEHAVIOUR OF SANDS AND CLAYS 132
4.5 CRITICAL STATE THEORY 137
4.6 PROBLEMS 150
4.7 REFERENCES 152

CHAPq\177R 5 Stress Distribution and Settlementin Soils

5.1 INTRODUCTION ! 53

5.2 FUNDAMENTAL EQUATIONS OF ELASTICITY 154

5.3 STRESS DISTRIBUTION DUE TO EXTERNAL AND 12VTERNAL LOADING ! 57
5.4 ELASTIC SETTLEMENT
OF FOOTINGS 184
5.5 SOIL-FOOTING INTERACTION MODELS 199
5.6 PROBLEMS 206
5.7 REFERENCES 210

CHAPTER 6 One Dimensional Consolidation 213

6.1 INTRODUCTION 213
6.2 CONSOLIDATION INDICES AND SETTLEMENT PREDICTION 213
6.3 SOLUTION OF ONE DIMENSIONAL CONSOLIDATION
DIFFERENTIAL EQUATION 226
6,4 APPLICATION OF PARABOLIC ISOCHRONES 243
6,5 LIMITATIONS
OF ONE DIMENSIONAL CONSOLIDATION THEORY 251
6,6 PROBLEMS 256
6,7 REFERENCES 258

CHAPTER 7 Application of Limit Analysis to Stability Problems

in Soil Mechanics 261

7.1 INTRODUCTION 261

7.2 LOWER BOUND SOLUTION 263
7.3 UPPER BOUND SOLUTION 280
7.4 FINITE ELEMENT FORMULATION
OF \177 BOUND THEOREMS 291
7.5 LIMIT EQUILIBRIUM
METHOD AND CONCLUDING REMARKS 298
7.6 PROBLEMS 299

7.7 REFERENCES 301)))

 Contents vii

CHAPTER 8 Lateral Earth Pressure and Retaining Walls 303

8.1 INTRODUCTION 303

8.2 EARTH PRESSURE AT-REST 303
8.3 RANK[NE'S THEORYFOR ACTIVE AND PASSIVE SOIL PRESSURES 305
8.4 COULOMB WEDGE ANALYS1S 3[4
8,5 COMMON TYPES OF RETAINING STRUCTURES AND FACTOR OF SAFETY 323
8.6 STATIC ANALYS1S OF CANTILEVER AND GRAVITY RETAINING WALLS 326
8.7 STATIC ANALYSIS OF SHEET PILE WALLS 331
8.8 INTERNALLY STABILIZED EARTH RETAINING WALL 354
8.9 THE OVERALL STABILITY OF RETAINING STRUCTURES 358
8.10 PROBLEMS 361
8.11 REFERENCES 364

CHAPTER9 Stability of Earth Slopes 367

9.1 INTRODUCTION 367

9.2 STABILITY OF SLOPES IN Cu, \177u- 0 SO1L

- C1RCULAR FAILURE SURFACE 368
9.3 STABILITY OF SLOPESIN c', O' SOIL - THE METHODOF SLICES 372
9.4 STABILITY OF INFINITELY LONGEARTH SLOPES 381
9.5 STABILITY OF REINFORCEDAND NAILED EARTH SLOPES 385
9.6 GENERAL SLOPE STABILITY ANALYSIS 393
9.7 APPLICATIONOF THE WEDGE METHOD TO UNREINFORCED SLOPES 399
9.8 CONCLUDING REMARKS 403
9.9 PROBLEMS 406
9.10 REFERENCES 408

CHAPTER 10 Bearing Capacity of Shallow Foundations and Piles 411

10.1 INTRODUCTION 411

10.2 ULT1MATE BEARING CAPACITY OF SHALLOW FOUNDATIONS 411

10.3 FIELD TESTS 425
10.4 AXIAL ULTIMATE BEARING CAPACITY OF PILES 433
10.5 PILE GROUPS 446
10.6 PROBLEMS 45O
10.7 REFERENCES 451

Index 455)))
Preface

Soil Mechanics: Basic Concepts and Engineering, Applications is primariIy designed as a main
text for university students taking first degree courses in civil engineering as well as
environmental and agricultural engineering. Emphasis is placed on presenting fundamental
behaviour before more advancedtopics are introduced. The speciaI structure of the book,
embodied in each chapter, makes it possible to be used in two, three and four year
undergraduate courses in soil mechanics. However, as new and advanced topics that extend
beyond standard undergraduate courses are included, the book will also be a valuable resource
for the practicing professional engineer. A problem solving approach is adopted through all
chapters and 152 worked examples demonstrate the engineering appIications, simulate
problem solving Ieaming and \177acilitate self-teaching. There are 113 unsolved problemswith
answers set for solution by students.
The use of S.I. units throughout, and frequent refer.ences to current international codes of
practice and refereed research papers, make the contents universally applicable. The main
feature of the book in comparison to a traditional soil mechanics text is the inclusion of the
soil reinforcement in retaining walls and slopes and tunnel stability problems.
Thisbook is written for readers that have no prior knowledge in soil mechanics, however a
reasonable background in engineering mechanics is essential.
I am most grateful for publishers, individuals and engineering organizations that have
permitted the use of their materials in this book. It must be noted that all of the refereed figures

are reconstructed and no copy has been carried out.

The author is grateful to the colleagues in the Faculty of Engineering and Surveying in
USQ for their continuous encouragements. Special thanks are due to Dr A. Kilpatrick of La

Trobe University, Australia for his help and time in proofreading of the material.
I am in debt to my family and my sincere thanks go to my wife Pari for her patience during
this project.
A. Aysen, M.Sc., Ph.D.
The University of Southern Queensland, Australia
[email protected],au
February 2002)))
CHAPTER 1

Nature of Soils, Plasticity and Compaction

1,1 INTRODUCTION

This chapter describes the basic characteristicsof soil, and physical properties that affect its

engineering behaviour. For a civil engineer the geological history of a soil may not seem to be
of much importance as most of the time a civil engineer is searching for a stress-strain model
justified by laboratory However some information
results. is necessary in order to facilitate
development of the model the physical
and understand and chemical behaviour of the material
of interest and, consequently, a brief description of soil chemistry is included. A significant
part of the chapter is devoted to the phase relationships that describe volume-mass relatedpa-
rameters which control the engineering behaviour of the soil. Essential information about par-
ticle size analysis, plasticity and soil classification is provided. Moredetails about laboratory
procedures and soil classificationmay be obtained from the standard codes of the relevant
country. The final part of the chapter discusses soil compaction and its importance to the soil.

1.2 NATURE AND CHEMISTRY OF SOILS

1.2.1 Origin of soils,geological

cfass\177cation and primary minerals

Soils are the unconsolidated layers that cover surface.

the For a civil engineer soil con-
earth's
sistsof particles of different size and shapeminor bonds between them
with forming a struc-
ture that undergoes deformation when subjected to natural or artificial forces. For an environ-
mental engineer soil is a product of the environment, and its formation is a function of climate,
organisms, parent rock or material, relief and time (Jenny, 1940). Soil mechanics is the art of

applying the mechanics of solidsto the soil to predict its deformation behaviour and strength.
This allows the design of soil structures and the investigation of its interaction with other
structures built on it.
Soils are derived from the weathering of parent rocks. Weathering or disintegration of rock
is of two types: physical and chemical. The physical weathering is the action of the forces as-
sociated with wind, water, glaciers and successive freezing and thawing. The chemical
agents
are water, oxygen, and carbon dioxide. In the physical weathering the mineralogical composi-
1)))
2 Soil Mechanics: Basic Concepts and Engineering Applications

tionof the parent rock does not change and the resulting soil inherits similar minerals. In the
chemical weathering the alteration of the minerals of the parent rock creates new minerals.
The processof soil formation is controlled by semi-independent factors of time, climate, or-
ganisms and topography. The time factor controls the equilibrium of the process with the envi-

ronment, a process that may take more than a million years (FitzPatrick, 1983).
The geological classification places rocks into three major groups, viz., igneous, metamor-
phic and
sedimentary rocks. Igneous rocks are those that have cooled down and crystallized
from a molten material called magma. of a variety
These rocks
of mineral crys- are composed
tals, the size of whichdependson the rate of cooling. The igneous rocks within the earth's
crust are formed by slow cooling and therefore have large crystals. They can be found on the

surface of the earth if the overlying material is removed or eroded.Igneous rocks formed on
the surface of the earth have been subjected to rapid cooling and therefore have small crystals.
Residual soils are a result of the physical and chemical weathering of these igneous rocks and
the constituent minerals are different to those of the parent rock. Sands have quartz and mus-
covite, similar to quartz of the parent
androckmuscovite
granite. The orthoclase feldspar and
pIagio\302\242lase of down into the clay minerals of kaolinite and montmorillonite.
granite break
Metamorphic rocks originate from residual soils under extreme temperature and pressure.
Both metamorphic rocks and residual soils are subjected to weathering and transportation
processes that result in the formation of sedimentary deposits. Compactionof these deposits,
together with cementation due to chemical reactions,produces sedimentary rocks. The sedi-
mentary rocks are also produced directly from metamorphic rocks within a weathering, trans-
portation, consolidation and cementation process. These sedimentary rocks could return to
their origin as either sedimentary deposits or metamorphic rocks as a result of weathering or
extreme temperature and pressure. This is an ongoing geological cycle through time and the
residual soils and sedimentary deposits so formed are known to us as soils. The soil particles
are of various sizes and are broadly categorized as clay, silt, sand, gravel and rocks. Each size
fraction exhibits different mechanical properties that can be successfully predicted using the
principles of soil mechanics.
Typical depositionalenvironments in soil formation (Fookes & Vaughan, 1986) include:
glacial environment, lake and marine environment and desert environment. For soils and soil

mechanics the most significant recent geological events beside the depositional environments
are rising and falling land and sea levels that result in further deposition and weathering. In a
geotechnical site investigation a geotechnicaI or civil engineer needs the cooperation of a ge-

ologist to identify the deposition environment and the subsequent deposition and weathering
cycle (Atkinson, 1993). The effect of mineralogical history on the mechanical properties is
evident in clay soils where the very smalI particles exhibit colloidal properties.
Minerals are divided into two groups
- primary and secondary- depending on their state in
the geological cycle. Primary minerals are those who inherit the properties of parent rock and
their chemical compositions remain unchanged. The common primary minerals in soils are
shown in Table 1.1 (Tan, 1994). The sand and silt fraction of the most soilsconsistslargely of
+ +
primary minerals. Quartz (SiO2) and feldspars (MAISi3Os, M = Na +, K and Ca cations) are
the most abundant primary minerals in soils (Bohn, et al., 1985). A rock of igneous origin may

have 50% to 70% quartz.)))

 Nature of Soils, Plasticity and Compaction 3
T,\177ble 1.1. Common primary minerals in soils.

Primary mineral Chemical composition

I. Quartz SiOz
2. Feldspar:
Orthoclase, microcline KAISi308
Albite(plagioclase) NaAISi308
3. Mica
Muscovite H2KA13Si3Ot2
Biotite (H,K)2(Mg, Fe)2(A1,Fe)2 Si3Oaz
4. Ferromzgnesians
Hornblende Ca,(Fe,Mg)z SiaOtz
Olivine (Fe,Mg)2 SiO,\177
5. Magnesium Silicate
Serpentine H,\177Mg3Si209
6. Phosphate

Apatite (Ca3(PO,0z)3.Ca(Fe,C])2
7. Carbonates
Ca/cite CaCO3
Dolomite CaMg(CO3)2

Quartz is believed to have been crystallized at low temperatures (Tan, 1994) and therefore it is
reasonably stable at ambient temperatures and resistant to weathering but the particles eventu-
ally become rounded in nature. Other primary minerals such as micas, and a variety of acces-
sory minerals of primary origin, are present in soils but in smaller quantities than quartz and
feldspars. Primary minerals may also occur in the clay fraction, but these are not the major
constituents of the clay soils. Secondary minerals are producedby weathering of rocks and
primary minerals and have different chemical composition from their parent materials. The
clay fraction of most clayey soils consists of secondary minerals.
The Jackson-Sherman weathering stages considersthree stages in the weathering process
(Sposito,1989).These are classified as early stage, intermediate stage and advanced stage, a
classification that indicates the intensity of leaching and oxidizing conditions. The products of
the early stage consist of carbonates, sulphates and primary silicates. Intermediate stage
weathering results in quartz, muscovite and secondary alum/no-silicates of clay fraction of soil
including mica and smectites. The advanced stage creates kaolinite that is an important clay
mineral. In general clay minerals fall into intermediate and advanced stages as shown in Table
1.2. This Table shows the increasing order of persistence of the soil minerals and their occur-
rencein time. The size of the mineral, its shape and its hardness controls the rate of weathering
and their engineering behaviour (Ollier, I972). The large minerals are more resistant to the

weathering because of their low specific surface which reduces the area of contact against the
weathering agents. The clay minerals have a fine texture, with individual particles less than
0.002 mm in diameter. As a result of the large surface area associatedwith the small particles
and masses, clay minerals develop plasticity when mixed with water. A knowledge of the
crystal structure of clay minerals is necessary in order to understand their behaviour.)))
4 Soil Mechanics: Basic Concepts and Engineering Applications

Table 1.2. Jackson-Sherman weathering stages (Sposito, 1989).

Clay minerals in soil fraction Soil chemical and physical conditions

Earl? stage
Gypsum Very low content of water and organic matter
Carbonates Very limited leaching
Olivine / pyroxene / amphibole Reducing environments
Fe-bearing micas Limited amount of time for weathering

Feldspar
Intermediate stage
Retention of Na, K, Ca, Mg, Fe(II),and silica:
Dioct\177dmdral mica / illite Ineffective leaching and alkalinity
Vermiculite / chlorite Igneous rock rich in Ca, Mg, Fe(ll), but no
Smectites Fe(II) oxides
Silicates easily hydrolysed
Flocculation of silica, transport of silica into
the weathering zone
Advanced stare
Kaoline Removal of Na, K, Ca, Mg, Fe(lI),and silica:
Gibsite Effective leaching, fresh water
Iron oxides (geotite, hematite) Oxidation of Fe(II),acidiccompounds, low pH
Titanium oxides (anatase, futile, ilmenite) Dispersion of silica, Al-hydroxy polymers

1.2.2 Crystal structure of clay minerals

In general, clay minerals are alumino-silicates
made by a combination of silica and aluminium
oxide units with metal ions substituted within the crystal. diagram of a silica unit
A schematic
is shown in Figure l.l(a). Each silicaunit consists ofa tetrahedronfour oxygen
with atoms, O,
located at the vertices of the tetrahedron equidistant from each other. A silicon atom, Si, is lo-
catedinsidethe tetrahedron equidistant from the oxygen atoms. These units combine to form a
silica sheet as shown in Figure 1. l(b). The arrangement could happen in variety of ways. Fig-
ure 1.1(b) shows an arrangement in a hexagonal pattern in which each basal oxygen atom is
shared with the adjacent unit. This sharing results a negative charge in the basic unit which can
be increased to zero if, for example, aluminium replaces silicon. Figure 1. l(c) shows the short
hand symbol generally used for silica. Figure 1.2(a) shows a unit of alumina (aluminium ox-
ide) where six hydroxyl ions surround one aluminium atom, A1, in an octahedral arrangement.
The combination of alumina units in staggered rows results in the formation of an alumina
sheet where each hydroxyl ion is shared by three basic units (Figure 1.2(b)). This sharing re-
suits a positive charge in each basic unit which can be reducedto zero if magnesium or iron
replaces atuminium. Figure 1.2(c) shows the short hand symbol generally used for alumina.
The basic elemental structure of a clay mineral is formed by different combinations of silica
and alumina sheets. Their type and engineering characteristics depend on the strength of the
bonds that connect these elemental structures.)))
 Nature of Soils, Plasticity and Compaction 5

-2

Silicon
\342\200\242

(a)

Short Hand symbol

(c)
Figure l. 1. (a) A basic silica unit, (b) silica sheet, (c) short hand symbol for silica sheet.

(a) (b)

Short Hand symbol

(c)
Figure 1.2. (a) A basic alumina unit, (b) alumina sheet, (c) short hand symbol for alumina sheet.

Kaolinite. Kaolinates, with the general chemical composition of 2SiO2A12032H20, are pro-
duced from weathering of the parent rocks that have orthoclase feldspar granite). An ele- (e.g.
mentary layer of kaolinite crystal is made of one silica sheet and one alumina sheet mutually
sharing the oxygen atoms between them. These Iayers join together to form a kaolinite particle
as shown in Figure 1.3. The number of elemental Iayers in one stack may reach one hundred or
more. The elementary layers are heId together with hydrogen bonding giving a stable structure

to the mineral, The hydrogen bonding is a result of the attraction forces between the oxygen
atoms of the silica sheet and the hydroxyl ions of the alumina sheet. These forces are high
enough to prevent the penetration of water between the layers and consequential expansion,
The mineral is moderately plastic when it is wetted and has some frictional resistance to shear.

Halloysite is another type of kaolinite mineral with a general chemical composition of

2SiO2A12034H20 and a tubular structure with water molecules between the elemental layers
of the kaolinite crystals.)))
6 Soil Mechanics: BasicConcepts and Engineering Applications

(a)

(b)
Figure 1.3. E|ementa[ structure of kaolinite.

As a resuIt, in contrast to kaoIinite, susceptible to expansionand contraction.

it is Upon heat-
ing, halloysite is irreversibly dehydrated and is converted to metahalloysite being a precursor
of kaolinite following the weathering sequence: Igneous rocks --> smectite(monmorillonite)
--* halloysite --> metahalloysite --> kaolinite (Tan, 1994).

llfite. Illite is producedfrom the weathering of micas with the major parent mineral of musco-
vite. A schematic representation of illite structure is shown in Figure 1.4. Illite has a mica type
structure in which an elementary layer is made from one alumina sheet sandwiched between
two silica sheets. The bond between the two elementary layers is made by potassium, K +,
which joins the six oxygen atoms of the two silica sheets. This bond is not as strong as in kao-
linite and, with a random staking of layers, there is more space to water to enter between the
elemental layers.

Montmorillonite. MontmoriIlonite, with chemical composition of 4SiO2AI203H20 + nH20, is

a product of the weathering of volcanic ash in marine water under poor drainage conditions.
Figure 1.5 shows the elemental structure of montmorillonite. Similar to illite the structure is of
mica type but the bond between the elemental layers is by weak oxygen-to-oxygen links com-

prises of water molecules and exchangeable cations.

s,

si
,,/\"

Figure 1.4. Elemental structure of illite.)))

 Nature of Soils, Plasticity aJ\177d Compaction 7

$i
-
(a) \\ H\177O

i S

Figure 1.5. Elemental stmclure of montmori[lonite.

The major characteristic of this mineral is its volume expansion or contraction due to an in-
crease or decreasein moisture content. At high moisture contents the mineraI is plastic and can
be deformed easilyunder minor stress but hardens when dry. The type of structural units and
their bonding influences the shape of the clay minerals and their surface area. As a result they
are either plate shaped or tubular. The average diameter varies from 0,1 to 4 lam, while the
-7
thickness has a range of 30 Angstrom (1 Angstrom = 10 mm) to 2 \177m. In general particle
size decreases from kaolinite to illite and to montmorillonite. The surface area for a constant

mass (specific surface) is a function of shape and size. The specificsurface of montmoriIlonite
is approximately 50 to 60 times that of kaolinite and 10 times that of illite. Due to the high ad-
sorption of water by montmorillonite, the mixture of this group of minerals and water, into the
state of a viscousfluid, is used in engineering construction as a grouting material to reduce
void ratio and permeability of soils (ASCE,1962).

1.2.3 Soil organic matter

Soil is a multiphase system containing solids, liquids, and gases in which the gas and liquid
phases are essentially air and water respectively. The solid phase consistsof inorganic and or-

ganic material. The organic material contains most of the carbon of the soil and usually con-
sists of both dead and living matter. Dead organic matter is the remains of dead plant material
and animal residues. It is the main source nutrients for plants and influences the engineering
properties of the soiI. The live microbial and plant matter comprises a few percent of the total
organic matter, the average amount of which may change up to 5% to 10% (by mass) depend-
ing on the drainage conditions, in some peat soils, however, the amount of organic matter ap-
proaches 100%. According to White (1979) the amount of organic matter on the surface of
soils (e.g. Australia) varies up to 16%, but decreases rapidly with depth. Organic soils and

peats are recognized by their high liquid limit (the amount of water needed to turn the soil into
a viscous liquid) but the range of moisture content in which the soils show plastic behaviour is
small. Organic soils, especiaIly those with a spongy structure, are also highly compressible
and do not meet the settlement limitations of most structures that may be built upon them.)))
8 Soil Mechanics: Basic Concepts and Engineering Applications

Table 1.3. Methods of tests for soils for civil engineering purposes: chemical and electro-chemicaltests.
Test ASTM BS AS AASHTO

Organic matter Loss on ignition: Titration: 1377 1289,4.1. I Titration:

content D-2974 Loss on ignition: 1377 T194
Sulphate Gravitational: 1377 1289.4,2. I
content Ion exchange: 1377
pH Value Electrometric: D- I067 Electrometric: 1377 1289.4.3. !
Colori metric: D-1067 Colorimetric:
1289.4.3.2
Electrical Disc electrodes:1377 1289.4.4.1
resistivity
Carbonate Titration: 1377
contenl
Chloride Water-soluble and
content Acid-soluble: 1377

The determination of the organic matter in a soil and its chemical composition - sulphates,
carbonates and chlorides - is recommendedin civil engineering construction works because of
their detrimental effects oll its engineering properties (Table 1.3).

1.2.4 Electric charge in soils

The solid phase of most soils carriesa net negative charge which is associated with the small
colloidalparticles of inorganic and organic origin. While the charge is negative on the surface
of the clay particle; the edges carry positive or negative charges due to broken bonds. The soil
particles that are responsible for cation exchange are referred to as colloid particles. A colloid
is a very fine particle of organic or inorganic origin with a maximum size of 0.2/am and a
minimum size of 50 A. A permanent charge is a result of isomorphous substitution in the crys-
+ + +
tal structure of the clay minerals (e.g. substitution of the AI3 ion for Si4 or Mg2 for AI3+).
The resulting charge can be neutralized by the cations that are positively charged atoms pre-
sent in the solution surrounding the clay particles. In illite and mica clays the potassium bond
is structural rather than exchangeable (McLaren & Cameron, 1996),and the cation exchange
capacity is relatively low. The unbalanced electric charge is responsiblefor the formation of
clay structure. Surface-to-surface contactis not possible as both have net negative charges
whilst the edge-to-surface contactis common as a result of the attraction between the positive
and negative charges. The electric charge does not affect the engineering behaviour of coarse
particles, as the ratio of surface to mass is low in comparison to that of the clay particles. A
variable charge is a resuIt of a change in pH in organic soils. A decrease or increase in pH will
result a decrease or increase ill negative charge, in some soils that have iron and aluminium
oxides, there are some areas, aIong the colloid particle surface, that have a positive charge and
wilI be neutraIized by the adsorption of negatively charged anions. The quantity of exchange-

able anions held is much smaller than the amount of exchangeable cations. The distribution of
cations and anions from the surface of the particle towards the voids created by the arrange-
ment of the particles has been modelled by different molecular adsorption models (Sposito,)))
 Nature of Soils, Plasticity and Compaction 9

1989).The simplest model is the Helmholtz double layer model in which it is assumed that the
layer balancing the negative charges is immediately adjacent to the surface colloid. In
of the
the diffuse double layer, referred to as the Gouy-Chapman model, the ion concentration de-
creasesgradually away from the surface of the colloid particle until it becomes equal to the ion
concentration of the bulk solution within the voids (Figure 1.6). The forces of attraction or re-

pulsion between electrical charges can be calculated mathematically from Coulomb's law.
Combining the equation of cation attraction with the mathematical expression of diffusion, a

relationship can be found between the electric potential and the distance from the colloid sur-
face (Boltzmann equation). Thus, the characteristics of the diffuse double layer may be pre-
sented by three types of distribution that include the distribution of cations (Figure 1,6), the
distribution of anions and the electric potential within the layer. The thickness of the diffuse
layer depends on the type of the ion and its bulk concentration and for natoral soils it may
range from 1 to 20 \303\227 10 ram. This model considers only the electrostatic forces of attraction
and repulsion and disregards the specific forces created by the finite size of the ions, The Stern
model is a modified version of the Gouy-Chapman model in which the double layer is divided
into two layers of the Stern layer immediately next to the colloid surface and the Gouy-
Chapman model beyond this layer (Figure 1.7). Electrical potential decreases linearly within
the Stern layer and exponentially across the diffuse layer.
The investigation of the nature of the layer of water surrounding a colloid particle is impor-
tant in soil chemistry, agriculture and environmental soil science from the view of exchange-
able cations available for plant intake. Furthermore, the clay fraction may intercept some pol-
lutants, carried by soil water, due to its cation exchange capacity and thus acts like a natural
purifying agent. This layer is strongly attracted to the soil particle and has different properties
from the water in the pore voids. In soil mechanics the study of the nature and mechanisms of
held and attracted water is important, as it is responsible for the plastic behaviour of the soil.
The term adsorbed water is used to describe this layer; whilst the water beyond the boundary
of adsorbedwater is called free water, The water particles within the adsorbed layer can move

parallel to the surface of the colloid particle, but the movement in the direction perpendicular
to the surface of particle is not possible or is restricted.

Thickness o! Gouy-
.\177
-.\177 Chapman \177ayer

Normal concentration
in pore water

Distance from colloid surface

Figure 1,6. The Gouy-Chapman double layer model.

L)))
l0 Soil Mechanics:BasicConcepts and Engineering Applications

Linear
\177
/[\177/, Exponential

\177 \177npore water

Distance from colloid surface

Figure 1.7. The Stem double layer model.

1.2.5 Soil structure

SoiI structure is the three dimensionaI geometrical arrangement of pores and particles of vari-
ous sizes. Some geoIogists have referred to the structure as soiI fabric to make it compatible
with the term rock fabric which describes the arrangement of mineral grains in the voIume of
the rock. Brewer (1964) defines fabric the physicaI constitution of a soil material
as as ex-

pressed by the spatial arrangement of thesolid particles and associated voids. The spatial ar-
rangement of coarse material is called the soil skeleton. Brewer & Sleeman (1988)suggest that
the soil structure is based on the arrangements of solids and pores in which the primary parti-
cles form a compound and the arrangement of compounds constitutes the soil structure. Al-
though in many definitions the emphasis is on the three-dimensional nature of the soil struc-
ture, two-dimensional studies of stability problems and soiI water movement have been very
successful in soil mechanics. In cIay soils the interparticle forces described previously have a
distinct effect on the arrangement of the clay particles. If the resultant of the forces creates a
net repulsion then a face-to-face arrangement called a dispersed structure will be created (Fig-
ure 1.8(a)). The net attraction results in an edge to face structure referred to as aflocculated
structure (Figure 1.8(b)). Experience indicates that clay deposits that are developed in fresh or
salty waters have a floccuIated structure, while the transported and remoulded cIaystend to
have a dispersed structure. The flocculatedstructure in salty water is more dominant than in
fresh water. In salty clays (e.g. marine clay) the concentration of cations is high, resulting in
thin adsorbed layers around the surface of the particles. The attraction forces dominate the par-
ticle arrangement; thus an edge to face structure is made. A clay soil with pure water moves
towards the dispersed structure. In laboratory techniques based on sedimentation, a salt, com-

monly sodium hexametaphosphate, is added to the suspension liquid as a dispersing agent to

facilitate particle sedimentation. Sometimes the aggregations made by original mineral units
make dispersed or floccuIated type structures depending on the deposition environment (Fig-
ures 1.8(c) and 1.8(d)). Theseaggregations have face-to-face structures, as there are little or no
attraction forces between mineral units. Depending on the velocity of the water in which the

particles settle these aggregations, or fine silts and sands, make a honeycomb structure as)))
 Nature of Soils, Plasticity and Compaction 11

shown in Figure 1.8(e). This structure contains a large amount of voids that are bridged by the
assemblage of aggregations or fine silts and sand particles. A diagrammatic reconstruction of
an electron microscope image of cIayey sand is shown in Figure 1.8(f) (Tan, 1994). The sand
particIes(S)form the main skeIeton of the soil structure.
Clay (C) and organic material (black
particIes) fiII the voids between the sands and voids are occupied by air. This
the remaining
reconstruction cannot be generalizedfor sands and coarse materials as the process of the struc-
tural arrangement of particIes is complex and must be assessed by standard tests and scanning
electron micrograpbs. An ideaIization of sand or a coarse-grained soil structure is shown in
Figures 1.8(g) and 1.8(h) for dense and Ioose structures respectively. The densemodel was
used by Rowe (1962) to demonstrate the stress-ditatancy in sand during shear.

1.2.6 Clay structural analysis: X-ray diffraction method

This technique was investigated

originally by Whittig (1964) and later suggested by Moore
(1970a, b) to the The X-ray diffraction method has probably
ASTM. contributed more to the
mineralogical characterization of the eIemental structure (soiI Iayer silicates) than any other
single method of analysis (Bohen, et aI., 1985). Optical and electron microscopy methods are

(d) (e)

(g) (h)

Figure 1.8. Different structures of clay, natural soiI and idealization of dense and loose sand.)))
12 Soil Mechanics: Basic Concepts and Engineering Applications

Figure 1.9. Geometric conditions for the X-ray reflection from mineral planes.

two other techniques used for the measurement of the fabric within small volume
a relatively

of soil. The optical microscopy is based on optical the

birefringence of the clay particles sub-
jected polarized
to light. In electron a
microscopy relatively larger area is scanned by the elec-
tron beam allowing the direct observation of the material at a wide range of magnifications. In
the X-ray diffraction method the material is exposed to a filtered X-ray beam as idealized by
Figure 1.9. The X-ray passes into the material and causes the electrons in the atoms of the
minerals to vibrate and reflect the beam through the successive planes. The method involves

increasing of incidence angle and monitoring the intensity of the diffracted X-radiation until a
maximum value of the diffracted intensity is achieved. The X-ray diffraction maximum is de-
tected whenever the following equation is satisfied:
nX = 2dsin0 (1.1)
where n = a series of integers, 1, 2, 3 ..... X =
known wavelength of the X-radiation, d = the re-
peated space between diffracting planes and 0 is the angle of incidence of the X-rays. Detec-
tion of the maximum diffracted radiation intensity is carried out either by a cylindrical photo-
graphic film placed around the sample or by a rotating detector. This technique is valuable for
measuring the thickness of the elemental structure as well as monitoring the interlayer spacing
with any change of the external conditions. The X-ray diffraction method is also used in soil
stabilization where the soil or a manufactured fine grained material such as fly ash is stabilized

by lime or cement. Development of hydration and bond formation between particles may be
monitored by this technique to justify the water content used for stabilization.

[.3 MASS-VOLUME RELATIONSHIPS

1.3.1 Basic physicalproperties

A soil sample contains soil particles (solids),liquid (usually water) and air. These three phases
can be visualized as three separate blocks as shown in Figure 1.10. This representation is an
artificial concept and all phases are inextricably mixed. The phase diagram is dimensioned in)))
 Nature of Soils, Plasticity and Compaction 13

Volume Mass

AIR

Vv
WATER

Vs SOLIDS

Figure I. I O. Dimensioned phase diagram.

terms of volume and mass. A completely dry soil or a fully saturated soil is a two-phase sys-
tem. Volume related symbols are defined as follows:
Va
= Volume of the air within the voids between particles.

Vw
= Volume of the water within the voids between particles.
Vs
= Volume of the solids.
=
V\177, Volume of the voids within a given sample = Va + V,..
V = Total volume of the soil sample = Vs + V\177.
Mass related symbols are: Ma =- Mass of air = 0, Mw = Mass of water, Ms = Massof solids.
M = Total mass of soil sample = Mw + Ms.
The common volume related parameters are as follows:

Void ratio e. The ratio of the volume of the voids to the volume of the solids:

e =Vv (1.2)

Specific volume v. The volume occupiedby a unit volume of solids:

V
- Vv +V\177
\342\200\242
v= , v=l+e (1.3)

Porosity n. The ratio of the volume of voids to the total volume:

v\177 e
,\177=v\177_\177,\177= = (\177.4)
v v\177 +vv
l+e

Void ratio is usually expressed as a decimal whilst porosity may be expressed as a percentage.

Degreeof saturation St, The ratio of the volume of the water to the volume of the voids:
Vw
Sr (L.5))))
14 Soil Mechanics: Basic Conceptsand Engineering Applications

Degree of saturation is usually expressed as a percentage. For a dry soil Sr = O. When the vol-
ume of the voids is fully occupied by water, Vw = Vv and Sr = 1 or 100%.

Air content Av. The ratio of the volume of air to the total volume:

(1.6)
V
The common mass-volume relatedparameters are:

Moisture content or water content m or w. The moisture content or the water content is the ra-
tio of the mass of water to the mass of the solids or dry mass of soil:

m or w=--Mw (1.7)
Ms
This parameter is usually expressed as a percentage.

Density of solids Os. The ratio of the mass of the solids to the volume of the solids:
Ms
p., = -- (1.8)
Vs
3= 3
1 g/cm =
3
In the SI
s2\177stem
the preferred unit is Mg/m3. Numerically 1 Mg/m 1 tonne/m =

1000 kg]mL

Specific gravity of solids Gs, The ratio of the density of the solids to the density of water:

Gs =
Ps
(1.9)
Pw
3
where Pw
=
density of water = 1 Mg/m at This
4\302\260C. is a dimensionless parameter within the

range of 2.6 to 2.8 depending on the mineralogy of the soil.

Density of soil 9. The ratio of the total mass to the total volume:

M
p = -- (1.10)
V

A preferred usage unit is Mg/m 3. The terms bulk density or wet density are also used.For a
given void ratio, this parameter changes from a minimum at the dry state to a maximum at the
saturated state:

M\177
= Ms +VvPw
Pd (Minimum), Psa\177 (Maximum) (i.11)
='--\177- V

The relationship between dry density, moisture content and bulk or wet density is:
p
Pal- (1.12)
l+w)))
 Nature of Soils, Plasticity and Compaction 15

Density index. To express the consistency states of sand and gravel, the natural void ratio is
compared with the minimum and maximum void ratios obtained in the laboratory. Density in-
dex (or relative density) is defined by:
- e
/D =
ero\177x
- (Ioosest) 0 _< 1/9 -< 1 (densest) (I.L3)
emax errdn

Useful equations:
Dry density expressed in terms of specific gravity and void ratio:

- GsPw
Pd (1.14)
l+e
Void ratio expressed in terms of moisture content, specific gravity and degree of saturation:
wGs
e = , e = wG s (for fully saturated soil) (1.15)
Sr
By combining Equations 1.14 and 1.15, a relationship is found to express dry density in terms
of moisture content, specific gravity and degree of saturation.

= GsPw
Pa (1.16)
1+ wG s/S r

- G s9\177\302\242
9d (for fully saturated soil) ('1.17)
1 \303\267
wG s

In a coordinate system with w on the horizontal axis and Pa on the vertical axis, this equation
represents the possible states of saturated soil and is termed zero air curve.

Dry density expressed in terms of specificgravity, moisture content, and air content.

Vw+V,\177 M,\177Ipw+VAv Mw+VAvp w GsM +GsVAvp w

was+GsAvPw
\177\" Ms fPs Ms fG.\177
M
s Pa

Substituting for e in Equation 1.14 and rea\177anging,

(1
- A
- Gs9 w v)
Pa (1.18)
1+ wG s

For zero air curve Av = 0 this \177uation reduces to Equation 1.17.

Combining Equation i. 16 and 1.18 the following relationship between degree of saturation, air

content and moisture content can be obtained.

-
- wG s (1 A
v)
Sr (1.19)
A\177+ wG s

Relationship between e, Gs,9w, and 9sa\177:)))

16 Soil Mechanics: Basic Concepts and Engineering Applications

- Gs + e
e= p\177,G s p\177,..\177
, or:\177sat =
\177w
-- (1.20)
- l+ e
\177sat \177
w

The submerg\177 or buoyant unit weight:

\177'
=
Ysaz (saturated unit weight) -\177w(unit weight of water) = (9saz - 9w)g,
using Equation 1.20:

Gs -1
Y'=Yw (I.21)
l+e

Example 1.1
A soil specimen has the following properties: Gs = 2.7, e = 0.6 and w = 14%. Calculate: dry

density, dry unit weight, wet density, wet unit weight and degree of saturation.

Solution:
When the void ratio e is known, we calculate the volume of solids Vs and the volume of voids
Vv in terms of a unit total volume (V= i m3).
3
V
v
+V
s =l,e=Vv/V orV
\177 s =l/(l+e),V v =e/(l+e),fore=O.6. Vs =1/(1+0.6) =0.625 m ,

V v = 1 - 0.625 = 0.375 m 3. Dry density is the mass of the solids that occupies a unit volume:
p\177
=
Gs9 wVs
= 1.69Mg/m
= 2.7 x 1.0x0.625 3.

= 16.6 kN/m 3.
= 1.69\303\2279.81
Dry unit weight: Yd

Wet density: 9 = Pa' (i + w)

= 1.69(1+ 0.14)= 1.93 Mg/m 3,

Wet unit weight: 7 = 1.93\303\2279.81 = 18.9 kN/m 3.

3
Vw
=
MwlOw =0,I4Ms/9w =0.237
=0-14\303\2271.69/l.0 m ,

Sr = V V = 0.237
w / v / 0.375 = 0.632= 63.2%.

Example 1.2

A soil sample has the following properties: Gs = 2.7, e = 0.7,Sr = 80%. Calculate: dry density,
dry unit weight, moisture content, wet density, wet unit weight and air content.

Solution:
3
For e = 0.70,V s = 1/(1 + 0.7) = 0.588m 3, Vt, = i - 0.588= 0.412
m .

Pd
= = =
GspwVs 2.7\303\2271.0\303\2270.588 1.59 Mg/m 3, y,\177 = 1.59\303\2279.81 = 15.6 kN/m 3.
3
S r =Vw/V v=0.8, Vw/0.412=0.8, V w =0.330 m 3, V
a =0.412-0.330=0.082 m.
Mw =
Vw \303\2279w
= 0.330\303\227 1.0 = 0.33 Mg, w = 0.33 / 1.59 = 0.207-= 20.7%.

P = Pd (1+ w) =
1.59(1 + 0.207) = 1.92Mg/m 3, A v= V
a /V = 0.082/1.0 = 0.082= 8.2%-

Example 1.3

In a field density test the following results are obtained:p = 1.95Mg/m 3, w

= 0.16. The spe-
cific gravity of solids is 2.7. Find:
dry density, void ratio, air content, degree of saturation,
saturated unit weight and moisture content at full saturation.)))
 Nature of Soifs, Plasticity and Compaction 17

Solution:

Od = 9/(1+ w) =
1.95/(1 + 0, 16) = 1,68Mg/m 3. Assume the total volume is 1 m 3, thus:
Vs.
=
9s = Ms/(Gs \303\227
Ms / 13 w)
= 1.68/(2.7 \303\227
1.0) = 0.622 m 3, V v = 1 - 0.622 = 0.378 m 3.

e =0.378/0.622 = = = -
0.608.V w M w / 9w (1.95 1.68)/1.0= 0.27 m 3,
Av =V a/V = (0.378-0.270)/1.0 = 0.108= 10.8%.
Sr =V w/V v =0.270/0.378 =0.714= 71.4%.
3
The volume of water at the saturated state = 0.378m and its mass is 0.378 Mg.Thus,
= 1.68+0.378 = 2.06 Mg/m3, Y\177r -- 2,06x9.81 = 20.2 kN/m 3.
9sa\177

Moisture content at full saturation = 0.378 / 1.68 = 0.225= 22.5%.

1.3.2 Experimental determination of physical properties of soils

Determination of related parameters require the direct
volume-mass measurement of the total
mass, mass of water, and specific gravity
total volume, of solids. The volume of a specimenof
an irregular shape is measured using water or mercury displacement techniques. Determina-
tion of the moisture content includes measurements of wet mass and oven-dried mass to estab-
lish the masses of the water and solid particles. In most standards the temperature of the oven
for drying purposes is 105\302\260Cto 110\302\260C,and the specimen must stay in the oven for 24 hours.

The specificgravity of solids. This parameter is measured using a simple apparatus called a
pyknometer which is a glass jar with a conical top as shown in Figure 1.11. The cone has a
hole at the top to alIow the water to overflow. procedure described in ASTM D-854,
The test
AASHTO T100, BS 1377and AS 1289.3.5,1, consists of the following steps:
1. Measure the mass of empty pyknometer (MI).
2. Fill the pyknometer with water until it overflows from the hole at the top of the cone and
find the total mass of the pyknpmeter and water (M:\177).
3. Empty the pyknometer and place a sample of oven-dried soil inside the pyknometer and
measure the combined mass (M3).
4. Add water until it overflows from the top of the cone and measure the total mass of the ap-
paratus (M,,). At this stage the jar must be agitated to remove the air bubbles. The specific
gravity is the ratio of the mass of dry soil to the mass of water displaced by soiI:
M 3 -M I
Gs = (1.22)
(M 2 -MI)-(M 4-M 3)

A similar method uses a narrow-necked density bottle (or volumetric flask) of 25, 50 or 100
ml volume that may be agitated by a mechanical device called an end over end shaker.

Field density. Common methods are based on excavating a hole at the site and measuring the
volume of the hole and the mass of the excavated soil. Field density is the ratio of the mass of
the excavated soil to the volume of the hole. The volume of the hole is measured in a variety
of ways using mostly sand or water replacement methods. In the sand replacement method two)))
18 Soil Mechanics: Basic Concepts and Engineering Applications

Figure 1.11. Pyknometer with rubber washers (Wykeham Farrance).

apparatus are available. In BS 1377 and BS 1924 the apparatus is an of a sand pour-
assembly
ing metal cylinder of 100 or 200 rmn diameter, a metal cone and tray of 300 ram
a metal

square with 100 or 200 mm diameter central hole (Figure 1.12(a)). The metal tray is adjusted
horizontally on the ground surface atad the soil is excavated through the hole in the tray to a

depth of approximately 1.5 times the diameter, with the resulting excavation being preferably
cylindrical in shape.

\177/L\177d

Sand

pouring
cylinder

(b)
\". _ .;\177\"- Excavated
hole

(a)

Figure t. 12.\177.a)Typical sand pouring apparatus, (b) calibration cylinder, (c) sand cone apparatus
(Wykeham Farrance).)))
 Nature of Soils, Plasticity and Cotnpaction 19

The sand-pouring cylinder is placed on the tray and fitted to the central hole. Opening the
shutter system allows the sand to flow through the cone, until the hole and t\177e cone are filled
with the sand. The mass of the sand in the excavated hole is calculated by measuring the mass
of the apparatus before and after the test, and knowing the mass of the sand in the cone. The
volume of the hole is equal to the mass of sand in the hole divided by the pouring density of
sand. To determine the mass of the sand in the cone the cylinder is filled with sufficient mate-
rial and the mass of the assembly is measured. The apparatus is sanded over the hole in the
tray which is placed on a flat and feveI surface. When the sand stops running the tap is closed
and the remaining mass of sand is measured. The difference between the two measurements
before and after the operation gives the mass of the sand in the cone. The pouring density is
measured by pouring the sand into a calibration cylinder (Figure 1.12(b)) o f known volume. A
similar method using a sand pouring can is recommended by AS 1289.5.3.2. The sand used is
cleanand dry, 90% passing a 1.18 mm sieve and 90% retained on a 600 I.tm sieve. Sometimes
an upper size of 2.36mm / 1.18 mm or a lower size of 600 I.tm / 300 I.tm is used. Care has to
be taken in shielding the sand to prevent moistening and bulking that may invalidate the re-
sults. The secondapparatus based on the sand replacement method is called sand cone appara-
tus (Figure t.12(c)) and is used by ASTM D-1556, AASHTO T191 and AS 1289.5.3.1 stan-
dard systems. This apparatus is used in holes where the maximum particle size does not
exceed 38 ram. The apparatus includes a plastic container of 5 litres capacity, a density plate

(tray) with 165 mm diameter central hole and a cone with 152 mm diameter that has a shutter
mechanism and threaded nozzle. According to ASTM D-1556-90(1996)recomrnendations,
the sand cone method is not suitable for organic, saturated, or highly plastic soils that would
deform or compressduring the excavation of the test hole. It may not be suitable for soils con-
sisting of unbounded granular materials or granular soils having high void ratios that will not
maintain stabIe sides in the test hole. The balloon density apparatus used by ASTM D-2167,
AASHTO T205and AS 1289.5,3.4, utilizes an empty rubber balloon to measure the volume of
excavated soil. The balloon is slowly filled with water from a calibratedglasscontainer until
the hole is completely filled by the expanded balloon. A direct water replacement method
(AS1289.5.3\1775) uses a density ring on the surface and a plastic sheet to retain the water in the
hole. This test is particularly appropriate for soils containing large particles as defined in AS
1289.0. In BS 1377 and BS 1924 the water displacement method is recommended only for
stabilized soils. In fine-grained soils the core cutting method is used to determine the fieId
density (ASTM D-2937, BS 1377and AS 1289.5.3.3). The core cutter commonly has a diame-
ter of 100mm and is 130 mm long and is driven into the soil using a driving ram of standard
weight. By measuring the mass of the soil in the core cutter and knowing its volume the field
density can be calculated.
Application of nuclear gauges for the determination of field density and in-situ moisture con-
tent provides a rapid method and moisture control (ASTM D-2922,D-2950,
for compaction
D-3017,BS1377and 1924). Readouts unitsin direct
are also provided for void ratio and de-
gree of compaction. A nuclear moisture-density gauge operates by emitting gamma and neu-
tron radiation from radioactive materials sealed within a capsule in the gauge. The gauge also
includes a detector of the G-eiger-Mueller type to collect and read gamma radiation emitted
from the cesium 137 and thereby determine the field density. For neutron radiation americium)))
20 Soil Mechanics: Basic Conceptsand Engineering Applications

(a) (b)

Figure 1.13. Transmission modes in nuclear moisture-density gauge.

241: berylium is used in the measurement of moisture content. A material of high density ab-
sorbs the gamma radiation and acts as a radioactive shield resulting in a low reading, whilst a

material of low density will give a high reading. The americium 241: berylium emits neutron
radiation to thetest material. Some of this radiation or high energy neutrons are moderated by
the hydrogen atoms of the water in the material and the rest is detected by the Geiger-Mueller
detector showing a reading on the gauge. A high reading in a specified period of time corre-
sponds to a wet material, while a low reading corresponds to a dry material. Based on the
amount of radiation detected in both cases the gauge could be calibrated either by the manu-
facturer or by the user. Two basic techniques, backscatter and direct transmission, are usedfor
the determination of moisture content and density as shown in Figure 1.13. Moisture content is
determined by using the backscatter mode only whilst density may be evaIuated by either
backscatter or direct transmission. The backscatter mode is shown in Figure 1.13(a) in which
both the radiation source and detector are on the surface of the test material. Directtransmis-
sion is not possible because of the intervening shield and the detector counts only on the radia-
tion received by reflection or radiation scattered back to the detector. Using this method depths
up to 65 mm may be analysed. In the direct transmission mode (Figure 1.13(b))a small hole is
drilled in the test material to a depth of 300 mm and, whilst the radioactive source is lowered
into the hole, readings are carried out at increments of 25 to 50 ram.

.t.4 PARTICLE SIZE DISTRIBUTION

1.4.1 Definition of soits and particle size distribution curve

Figure 1.14 shows a generally used size classification for soils as proposedby MIT (Massa-
chusetts Institute of Technology). Each group is divided into three sub-divisions of fine, me-
dium and coarse particles. Sands and gravels are coarse-grained soils; and considered as non-
cohesive (or cohesionless)materials. They are also referred to as granular materials. Clay and
silt are fine-grained and are classified as cohesive materials. I.n practice fine-grained materials
are soilspassing # 200 sieve, which has square openings of 0.075ram.)))
 Nature of Soils, Plasticity and Compactfon 21

0.002 0.06 2.0 60.0

CLAY S\177LT SAND GRAVEL COBBLES
0.006 0.02 0.2 0.6 6.0 20.0
F M C F M
[ I I 1C

Particle Size (mm)

Figure 1.14. Definition of soils according to MIT.

A particle size distribution curve describes the percentage by mass of particles of the different
size ranges, The horizontal axis represents the particle size on a logarithmic scale. The vertical
axis represents the percentage by weight of particles that are finer than a specific size on the
horizontal An example of this curve is shown in Figure
axis. 1.15. Consider point A on this
curve with coordinates of 9.5 mm and 82%. This means that 82% by weight of particles are
finer than 9.5 ram. In a typical particle size distribution test, sieves are stacked in a motorized
shaker in order of decreasing aperture size and a pan at the bottom of the sieve column retains
fine material that passes through all the sieves. For the tnaterial passing through sieve # 200, a
sedimentation test is carried out. Figure 1.16 shows three size distribution curves: soi\177 W is re-
garded as a wel\177-graded soil, soil U is a uniformly graded soil with a size range much less than
soil W, soil P is a poorly graded soil and lacks a particular range of particles.To describe the
state of grading, two parameters are defined as follows:

Coefficient of Uniformity: CU = D60/ DI0 (1,23)

Coefficient of Curvature: C c = D\177O /( D6oDIo ) (1.24)

where DI0, D30and D60 are particle sizes corresponding to 10%, 30% and 60% finer respec-
tively and are determined from the particle size distribution curve obtained in the laboratory.

tO0
9o II I1
I
80
70
6O I I11111

50 J/Irq]l] IIIIIII

40 111111 IIIIIII

20

0
0.001 0.01 0.1 1 10 100
Particle size (ram)

Figure 1.15. Particle size distribution curve. (Exm-nple 1.4).)))

22 Soil Mechanics: Basic Concepts and Engineering Applications

lOO
90 lacks this S
80 s'ze range
70 k P/ \177
60
50
40
30
20 [ / /
range only

10
0
0.001 0.01 0.1 1 10 1 O0

Particlesize(mm)

Figure 1.16. Well-graded, uniformly graded and poorly graded soils.

The magnitude of Cu indicatesthe spread of grain size values; a larger range leads to larger
number. According to the Unified Soil CIassificatlon, for well-graded soils,the vaIue of CC is
between 1 and 3. For a well-graded gravel, or gravel-sand mixtures with littIe or no fines, the
value of Cu is greater than 4. For well-graded sands, and gravelly sands with little or no fines,
the value of Cu is greater than 6.

1.4.2 Sieve and hydrometer analyses

The standard procedure for determining the distribution of particle sizes may be found in
ASTM D-422, BS 1377and AS 1289.3.6.1. Sieve requirements are specifiedin ASTM Ell,
BS 410 and AS 1152 and are manufactured in three diameters of 200 ram, 300 ram, and 450
ram. Normally, sieves have a mesh of woven brass wire with square apertures ranging from
0.075 mm to 125 ram. The sieve with 0.075 mm apertures (or sieve # 200, indicating 200

openings in 25.4 mm length) is the smallest practical sieve sizeused in soil mechanics labora-
tories. Prior to sieving, soil samples must be oven-driedat t05\302\260Cto 110\302\260C.For intermediate
and fine size particles the soil sample is first submerged in a dispersing solution at least for an
hour. It is then washed with water on the 0.075 mm sieveand oven dried. An analysis without
washing is also permitted by standard codes, but the results are not reliable if an appreciable
amount of clay and silt is present in the sample. Sieving may be done by hand or by a sieving
machine. A typical motorized shaker or sieving machine is activated by electromagnetic im-
pulses with a control panel for time adjustment and vibrating intensity. Some advanced shak-
ers permit the adjustment of pause time betweenvibrations. A sieve shaker holds up to ten
nesting sieves of any diameter. The particle size distribution for the fraction of the soll passing
through the 0.075 mm sieve (clay and silt) is carried out using the hydrometer method speci-
fied in ASTM D-422, AASHTO T88, BS 1377and AS 1289.3.6.2. The methods described in
these standard codes are not applicable if less than t0% of the soil passes the 0.075 tam sieve.)))
 Nature of Soifs, Plasticity and Compaction 23

The hydrometer is based on the measurement

method of the velocity of soil particles in a
se\177limentation the dry mass of soil in the solution
solution m\177d in different intervals of time.
The velocity of falling particles and the d\177 mass of soil at a specific depth are measured by a
hydrometer originally manufactured to measure the specific gravity of a solution. The results
are combined with the Stokes' law, which gives the relationship between the velocity of a
spherical particle and its diameter while settling within a solution. As a result, two relation-
ships are constructed between the hydrometer reading and both the diameter and percentage
finer than that diameter for each reading at a specific time. Stokes' law was originally pre-
sented in the following form:

where D is
dyne-see/era
the
2 ....
of a spherical particle in cm rl is the dynamic viscosity of water
diameter
umt weight of the partMe
is the in g-force/era
3
,
. .
Yw is the untt weight of water
in '
,\177Ys
in g-force/era \177,and v is the velocity of the spherical particle in cm/sec. The reading R of a hy-

drometer at the time t gives the amount of soil particles in gtl000 ml of solution (or equivalent
suspension density) at the centre of the hydrometer volume. The distance of a reading from
this centre, L (Figure 1.17(b)),represents the length covered by a particle of a specificdiame-
ter during this time. Replacing v with L/t in Equation 1.25:

O = \177]18a3L/t(y s
-
Yw) (1.26)

Equation 1.26 may be simplified using familiar SI milts:

D = L-\177'\177xl.749@t/(ps-p\177.) xl0/-J\177- (1.27)
3 2
where D is the diameter soil particle in
of the \177tm, Os and Pw are in Mg/m , rl is in 103 x N/m

x seconds and t is in minutes. The effective length L is found by calibrating the hydrometer.
Equation 1.27 canbe written in the following form:
D = F1 xF 2 \303\227
F3 (1.28)

Parameters Fl and F3 are calculated from the test data. Parameter F2 depends on the properties
soil and
of the water and can be tabulated for a specified Ps (Table 1.4).
2
A sample calculation for T= 15\302\260C:q = 0.001139 N/m x s, pw= 0.9991Mg/m 3,

F2 = - =
1.749\177/1.139/(2.7 0.999l) 1.432.
The percentage finer than D is calculated from the followiug equation:

Percentage finer = R
c x a / M s x 100% (1.29)
where M\177 is the mass of dry soil and Rc is the corrected value of hydrometer reading for tem-

perature, density of water, and meniscus effects. The parameter a corrects the reading due to
the specific gravity of solids as the hydrometer is made (e.g. hydrometer 152H, ASTM) to
measure the density of a solution in g/1000ml for Gs = 2.65:
0.623G
a =
1.65G\177 - s
(1.30)
2.65(G
s
-
i) Gs - t)))
24 Soil Mecfvanics: Basic Concepts and Engineering Applications

Table 1.4. Values of parameter F2 versus temperature for 9s = 2.7Mg/m 3.

Temp.\302\260C F2 Temp.\302\260C F2 Temp.\302\260C F2

10 1.534 17 1.395 24 1,282

11 1.512 18 1.378 25 1.267
12 1,491 19 1.361 26 1.253
13 1.471 20 1.344 27 1,239
14 1,451 21 t.328 28 1.225
15 1.432 22 1,312 29 1.212
I6 1,414 23 1.297 30 1.t99

The amount of oven-dried soil used for the test is 50 g in a clayey soil to 100 g in a silty soil
and the sample may be collected from the sieve washings and then oven-dried. The test is car-
tied out in i000 ml sedimentation glass cylinders (Figure 1.17(a)). The soil is mixed with wa-
ter and a dispersing solution (cormnonly sodium hexametaphosphate) and is shaken manually
for about one minute. The cylinder is then placed on a firm level and vibration free surface and
by immersing the hydrometer in the suspension the reading starts in logarithmic sequences of
time, An alternative method of obtaining suspension density is the use of an Andreasen pipette
(BS 1377). The 10 ml capacity glass pipette is used to obtain a small quantity of suspension
from a prescribed depth at different intervals of time to calculate the suspension density. The
pipette is fixed to a carriage assembly on a vertical stand and operates with lEO vibration whilst
the pipette is inserted and withdrawn from the liquid suspension. The cylinder used in this
method is 500 ml and samples are taken from a 100 mm depth.

Example 1.4

A sieve analysis on a soil sample of 978 g has given the following results. Plot the particle size
distribution curve, determine Co and Cc and classify the soil.

(a) (b)

Figure 1.17. (a) Hydromeler and sedimentation cylinder (Wykeham Farrance), (b) effective depth.)))
 Nature of Soils, Plasticity and Compaction 25

Sieve size (ram) Mass retained (g) Sievesize(nun) Mass retained (g)

75 0 1.18 118.5
37.5 25.5 0.6 83.1
19 52.8 0.3 39.9
9,5 97.8 0.15 24.6
4,75 207.3 0.075 27.3
2.36 298.2 Pan 3.0
Solution:
The results are tabulated below and shown irt Figure 1.15, from which: DIn = 0.63 ram, D30=
2.40mm, D60 = 4.57 nun. C =
U
=
D60 / D10 4.57/0.63
= 7.25,
C = 2.0. The soi\177 is-\177 with sand.
c Z\177o/(C'\1770Z\177,o)-- 2.402/(4.57 \303\2270.63)= we\177-graded grave\177

Sieve size (ram) Retained (%) Total retained (%) Finer than (%)

75 0.00 0.000 100.000

37.5 2.607 2,607 97.393
19 5.398 8.005 91.995
9,5 10.000 18.005 81.995
4.75 21.196 39,201 60.799
2.36 30.490 69.691 30.309
1.18 I2.116 81.807 18,193
0.6 8.496 90.303 9.697
0.3 4.079 94.382 5,618
0.15 2.515 96.897 3.103
0.075 2.791 99.688 0.312
Pan 0.306

Example 1,5

For the following data with Ms = 50 g and Gs = 2.7 draw the particle size distribution curve.
The hydrometer used is a 152H ASTM, where L = 163- R 49
\303\227 / 30 (mm).

T (Ca) Time (rain.) Reading T (C\302\260)Time (rain.) Reading

18 1 45,5 18 30 27.5
18 2 41.5 20 60 26.0
18 3 38.0 22 120 25.4
18 4 35.5 24 240 25.0
18 8 32.0 19 1440 24.0
18 16 29.0

Solution:
The results are tabulated and shown ila Figure 1,18. Sample calculation at t = 30 rain.:
L =i63.0-27.5x49/30= 118.1ram, Fi =-,illS. l/10 =3,436,
F2 = 1.378,F3 = 10/qcf\177.-.-\177 = 1.826, D = 3.436xl.378xl.826 = 8.64\177tm. From Equation 1.29:
(%) finer = 27.5 / 50 = 0.55 = 55%.)))
26 Soil Mechanics: Basic Concepts and Engineering Applications

\177_ \17700

N 80

\177
60

. 40
c 20
\177. 0

1 10 100

Particle size (#m)

Figure 1.18, Example 1.5.

T (C Q) Time (rain.) Reading L (mm) F1 F2 F3 D (ktm) (%) Finer

18 1 45.5 88.7 2.978 1.378 10.00 41.03 91.0

18 2 41.5 95.2 3.085 1.378 7.071 30.06 83.0
18 3 38.0 100.9 3.176 1,378 5,773 25.26 76.0
18 4 35.5 105.0 3,240 [.378 5.000 22.32 71,0
18 8 32.0 110.7 3,327 1.378 3.535 16.21 64.0
18 16 29.0 115.6 3,400 1.378 2.500 1 t.71 58,0
18 30 27.5 118.1 3.436 1.378 1,826 8.64 55.0
20 60 26.0 120.5 3.471 1,344 1.291 6.02 52.0
22 120 25.4 121,5 3.486 1.3t2 0.9[3 4.18 50.8
24 240 25.0 122.2 3.496 [.282 0.645 2+89 50.0
19 1440 24.0 123.8 3.518 1.361 0.263 1.26 48.0

1,5 INDEX PROPERTIES AND VOLUME CHANGE IN FINE GRAINED SOILS

1.5,1 Consistency states of fi'ne grained soils

The volume change and flow behaviour of a fine grained soil both depend upon its moisture
content, At a high level of moisture the soil has the whilst at a low mois-
properties of a liquid,
ture content it takes on the properties of a solid. At moisture contents between these two states
the soil passes from a plastic state to a semi-solid state as the moisture content decreases. The
physical condition of soil-water mixture is denoted as its consistency. Consistency is the resis-
tance to flow, which is related to the force of attraction between particles and is easier to feel
physically than to describe quantitatively (Yong & Warkentin, 1966). Figure 1.19 shows the
different consistency states of a mixture of water and fine-grained soil. The boundaries of
these states, expressed in terms of moisture content, are termed the Atterberg limits (after Al-
bert Atterberg, a Swedish agricultural scientist), and are defined as lbitows:)))
 Nature of Soits, Plasticity and Compaction 27

SL PL LL
I I
I I
\177 \177
Solid State Semi Solid State , Plastic State Liquid
\177 State
I I

Moisture Content

Figure [. 19.Consistency states.

Liquid Limit: The liquid limit LL is the moisture content above which the soil-water mixture
passes to a liquid state. At this state the mixture behaves like a viscous fluid and flows under
its own weight. Below this moisture content, the mixture is in a plastic state. Any change in
moisture content either side of the LL produces a change in volume of the soil.

Plastic Limit: The plastic limit PL is the moisture content above which the soil-water mixture
passes to a plastic state. At this state the mixture is deformed to any shape under minor pres-
sure.Below this moisture content, the mixture is in semi-solid state. Any change in moisture
content at either side of the PL producesa change in volume of the soil.

Shrinkage Limit: The shrinkage limit SL is the moisture content above which the mixture of
soil and water passes to a semi-solidstate. content, Betow this moisture the mixture is in a
solid state. Any content is associated with volume
increase in moisture change but a decrease
in moisture content does not cause volume change. This is the minimum moisture content that
causes full saturation of the soil-water mixture. The volume remains constant as the mixture
goes through the dry state to the SL moving from zero saturation to 100% saturation (Figure
1.20). On the wet side of the SL the volume of the mixture increases linearly with increasing
moisture content. On this liqe the mixture is fully saturated. A decrease in moisture content
moves the state of the mixture along the broken line CBA. Using these limits the following
dices are defined and used in the classification and description of fine grained-soils:

Plasticity Index PI = LL - PL (mixture is plastic at this range) (1.31)

=
Activity PI/Clay particles (%) (1.32)
Liquidity Index IL = (w- PL)/PI (w
= moisture content in the field) (1.33)
The Atterberg limits are used extensively of fine-grained soils. Finding
in the classification re-

lationships between these limits properties of the soil has been a matter of
and the engineering
researchfor many years. Early researchers such as Terzaghi & Peck (1967)suggested the di-
rect proportion of liquid limit and soil compressibility. Sherard (1953) reported similar behav-
iour while investigating the effects of index properties on the performance of earth dams.
Whyte (1982) suggested a method based on extrusion for the determination of the plastic limit
and found that the ratio of strength at the plastic limit to the strength at the liquid limit is ap-

proximately 70. According to Skempton & Northy (1953), however, this ratio is approxi-
mately 1,00. A comprehensive collection of equations relating compressibility indices and soil
plasticity is reported by Bowles (1996). These relationships may be useful in guiding the engi-
neer in the early stages of a feasibility study before conducting exteusive soil exploration and)))
28 Soil Mechanics: Basic Concepts and Engineering Applications

I
I
!
I
I
I

0 SL Moisture Content

Figure I.\1770. Volume-moislure content relalionship.

strength tests. The liquid limit for clay minerals may vary from 50% for kaolinite to 60% for
illite and up to 700% for montmorillonite. Kaolinite and illite may show moderate plastic lim-
its of 25% to 35%, whilst in montmorillonite the plastic limit can reach 100%. Note that ex-
changeable cations such as Na and Ca may influence clay activity, and this has to be investi-
gated if these minerals are to be subjectedto salty water.

1_5.2 Determination of the liquid, plastic and shrinkage limits of soils

77\177e Casagrande method for determination of the liquid limit. The relevant apparatus is shown
in Figure 1.21 and complies with ASTM D-4318, AASHTO T89,BS 1377, AS 1289.3.1.1 and
AS 1289.3.1.2. It is comprised of a brass bowl (A in Figure 1.21) that is hinged to a crank,
which, on rotation, causes the bowl to be lifted and dropped 10 1rim onto a hard rubber base -
a counter records the number of rotations (blows). A pat of soil is placed in the cup covering
approximately 2/3 of the area, and is grooved into two pieces with a standard grooving tool

(B). The cup is a part of a 54 rmn diameter sphere and has a maximum depth of 27 ram. The

original grooving tool was later replaced by the ASTM grooving tool (C) to overcome difficul-

ties in grooving silty and sandy soils. The test specimen is made from 250 g of soil passing the
425 gm sieve. Water is added in increments and mixed with the soil until the soil becomesa
thick homogeneous paste. It is then left at room temperature to cure for 12 hours. A part of this
mixture is placed in the Casagrande cup and levelled parallel to the base to make a depth of
approximately 10 ram. After closure of 10 mm of the groove as a result of the lifting and
dropping of the cup, the number of blows is recorded and a sample of the mixture is taken

from the centre of the closed groove for moisture content determination. The moisture content
of the remaining mixture is increased or decreased (by air drying) and the test is repeated three
times to give four different moisture contents and corresponding numbers of blows. These re-
suits are plotted on a semi-logarithmic scale with the number of blows on the horizontal loga-
rithmic axis and moisture content on the vertical axis on the linear scale. A line of best fit is
drawn and the moisture content corresponding to 25 blows is the liquid limit of the soil. Most
of the standard codes recommend a one-point test to establishthe liquid limit as a subsidiary
method accordingto:)))
 Nature of Soils, Plasticity and Compaction 29

Figure 1.21. Liquid limit device with grooving tools (Wykeham Farrance)\303\267

tan
LL = w (N / 25)
13
( 1.34)
N

where LL is the liquid limit in percentage, WN is the moisture content in percent corresponding
to N number of drops,and 13 is the slope of the semi-log plot.

The cone liquid lbnit of a soil. The cone penetration test (Figure 1.22) is described in BS I377
and AS I289.3.9 (allowedin I991). In this method the liquid limit is the moisture content of a

soil-water mixture placed in a standard cylinder when a standard cone penetrates 20 man into
\302\260
the soil paste in five seconds after it is released freely from the surface of the sample. The 60

Figure 1.22. Cone penetrometer wilh electric timer unit (Wykeham Farrance).)))
30 Soil Mechanics: Basic Concepts and Engineering Applications

Figure t.23, Shrinkage dish, glass cup and prong plate (Wykeham Farrance).

cone is manufactured from metal and has a height of 32 mm whilst the cylinder has a diameter
of 53 mm and a height of 40 mm. The procedure for sample preparation is similar to the
Casagrande method and the test has to be repeated at least four times using the original cured
sample.Results are plotted in log-linear scale, with moisture contents on linear vertical axis
and cone penetrations on logarithmic horizontal axis. A line of best fit is drawn through the
data and the moisture content corresponding to a cone penetration of 20 mm is determined.

Determination of the plastic limit. The test procedure is described in ASTM D-4318,
AASHTO T90, BS 1377and AS 1289.3.2.1 and is performed on the material prepared for liq-
uid limit test. The plastic limit is the moisture content at which a 3 mm diameter thread of soil-
water paste shears when rolled on a glass plate with the tip of fingers, and the test is repeated
several times to obtain an average value for the PL.

Determination of the shrinkage limit, This test is carried out according to ASTM D-427,
AASHTO T92 and BS 137% The procedure involves the measurements of mass and volume
of the soil-water sample (cake) in two states of wet, close to the liquid limit (point C in Figure
1.20) and oven-dried (point A in Figure 1.20). The soil cake is preparedin multiple layers in a
shrinkage dish of 45 mm diameter by 22 mm depth (dish A in Figure 1.23) and, after air-
drying, is left in the oven until completely dry. After measuring the mass of the oven-dried soil
cake it is submerged in mercury (dish B in Figure 1.23) and its volume is measured by mer-
cury replacement. A prong plate C is necessary to level the mercury and keep the soil cake
submerged in the dish. The volume of the shrinkage dish A is equal to the volume of the wet
soil and can be measured by mercury replacement method. By knowing the mass and volume
in two states of wet and dry the shrinkage limit can be calculated(Example 1.8). As the vol-
ume of the replaced mercury is measured manually, the user has to abide by appropriate safety
and health practices. Whilst shrinkage limit is not used in soil classification systems, its de-
termination is recommended to evaluate the swell potential of clayey soils. Alternatively, a

shrinkage limit apparatus, which accormnodates a cylindrical sample of 38 mm diameter by 76)))

 Nature of Soits, Plasticity' and Compaction 31

Figure 1.24. Shrinkage limit apparatus (Wykeham Farrance).

mm long, may be used for volumetric shrinkage determination. The apparatus includes a pre-
cisionmicrometer and both the volume and the volume change are measured by mercury re-
placement. A typical apparatus is shown in Figure 1.24. For the soils with low clay contents
where the plastic limit test is
to perform, a linear shrinkage test is recormnended (BS
difficult

1377,AS 1289.3.4.1). shrinkage moulds

The are semi-cylindrical troughs, 250 finn long with
an internal diameter of 25 ram, made from copper, plated steel, stainlesssteelor brass.They
are filled with a soil-water paste, having a moisture content close or ideally equal to liquid
limit. After 24 hours curing in a controlled temperature room the samples are oven-dried and
the mean longitudinal shrinkage LS is measured to the nearest millimetre. The percentage lin-
ear shrinkage LS of the specimen is defined by:

LS = (Ls I L)\303\227[00% (1.35)

where L is the of the mould in ram. If the quantity

length of materiaI used is limited, shorter
moulds (minimum 100 ram) with a semi-circular section can be used. In ASTM D-4943 two
shrinkage factors, volumetric shrinkage and shrinkage ratio, are defined as follows.

Volumetric shrinkage is the ratio of the volume change (due to moisture content change from
the liquid limit to the shrinkage limit) to the volume at the liquid limit:

SV =
(VLL -VsL)/VLL (1.36)

Shrinkage ratio is defined as the ratio of volume change (expressed as a percentage of volume
of dry soil) content change from limit to the shrinkage

....
to the moisture the liquid limit:

AV/Vdry AV/Vdry AV/Vdry Ms Od

\177 --G
SR =
(1.37)
Aw
AMw/Ms AVxf\177w/Ms VdryXOw 9w)))
32 Soil Mechanics: Basic Concepts and Engineering Applications

9O

\177
87

'- 85 ,<
o
o g4

\177 83

80
79
I0 25 100
Number of blows

Figure 1.25. Example 1.6.

Which is the apparent specific gravity of the soil in the dry state. Note that in the calcuIation of
the shrinkage limit and shrinkage ratio the initial moisture content does not affect the results.

Example 1.6
The following data were obtained from a liquid limit test using the Casagrande apparatus. De-
termine the liquid limit of the soil.

Number of blows Moisture content (%) Number of blows Moisture content (%)

11 89.1 32 81.8
20 86.2 45 80.1

Solution:
The results are plotted in Figure 1.25. The Iine of best fit is drawn and the liquid limit is taken
as the moisture content corresponding to 25 blows.Hence,LL = 83.8%.
Exarnple 1.7

Using the fo]Iowing data determine the cone liquid limit of the soiI.

Cone penetration (ram) Moisture content (%) Cone penetration (ram) Moisture content (%)

11.3 63.08 21.9 70.1

14.8 65.3 25.2 71.0

Solution:

Figure 1.26 shows the plot of moisture content versus penetration depth on a logarithmic scale.
The Iiquid Iimit is taken as the moisture content corresponding to a cone penetration of 20
ram. Hence, LL = 68.9%.)))
 Nature of Soils, Ptasticfty and Compaction 33

10 20 1O0
Cone penetration (ram)

Figure 1,26. Example 1,7.

Example 1.8
A fuIly saturated sample of soil in its natural state was found to have a total mass of 204 g and
a total voIume of 122 mI. When the sample had been oven dried the mass was 130 g and the
total volume 58 ml, Calculate the vaIue of shrinkage limit and Gs.

Solution:
Mass of water state (point C in Figure
at natural 1.20) = 204.0
- 130.0= 74.0 g.
3
VoIume of water at natural state = 74.0/l(g/cm 3) = 74.0cm = 74.0 ml.

Volume change (Point C to Point A in Figure 1,20) = 122.0 - 58.0= 64.0ml.

Volume of water at shrinkage limit (point B in Figure 1.20) = 74.0- 64.0= 10 ml.

Mass of water at shrinkage limit = 10.0

g.
SL = moisture content at shrinkage Iimit = 10.0/ 130.0= 0.769= 7.7%,
V\177=58.0
- 10.0 =48.0 mI, Gs= 130g/(48 ml x I g/ml) =2.71.

1.6 SOILCLASSIFICATION FOR GEOTECHNICAL PURPOSES

1.6.1 Introduction to classification of soits

Soil classificationfor engineeringis necessary to describe

purposes the many types of the soil
that exist in nature. system is a standard
A classification language (Craig, 1997) which organ-
ises the engineering knowledge of soils and is a means of communication. This sectionde-
scribesthe Unified Soil Classification System (ASTM D-2487)which is the basis for the AS
1726-1993. This Austratian standard does not preclude the usage of an alternative system and)))
34 Soil Mectvanics: Basic Concepts and Engineering Applications

consequently references are made to BS 1377-2:1990 where appropriate. The usageof an al-
ternative system is useful, as the Unified Soil Classification System may not adequately de-
scribe certain types of soils such as calcareous soils, pedocretes, and laterites. The classifica-
tion of Australian soils and the basic concepts of soil survey are given by Isbell (1996).
Most classificationsystems in soil mechanics use particle size characteristics, liquid limit
and plasticity index to describe the soil and its name. An engineering soil classificationsystem
is only useful for qualitative applications. In the construction of important soil structures the
classification must be supplemented by laboratory tests other than those needed for classifica-
tion. The assessment is made using disturbed samples recovered from site as well as undis-
turbed samples from boreholes and excavations. The particle size distribution is carried out on
material passing the 75 nun sieve in ASTM D-2487 and 63 rnm sieve in AS 1726. Index prop-
erty tests are performed on the material passing the 425 p,m sieve. More often a geological in-
vestigation is needed to identify the age and type of the deposition and its environment. The
soil description (using any standard system) should (a) distinguish between composition, con-
dition and of the soil, (b) describethe
structure information which may be obtained from a
disturbed soil and (c) describe the additional
sample, condition and structure properties which
may only be observed in an undisturbed soil (AS 1726-1993).

1.6,2 Unified Soil Classification System

The Unified Soil System evolved from the airfield
Classification classification system devel-
oped by Casagrande (1948). In this system two letters describe each soil. The first letter indi-
cates the dominant size whilst the second letter describes the fines, particle size distribution,
and plasticity of the soil (Table 1.5).Soils are divided into two major groups of coarse and
fine-grained material. Coarse-grainedsoilsinclude gravels, sands, non-plastic silts and their
mixtures with more than 50% larger than the 0.075 mm sieve. Fine-grained soils include silts,
clays and their mixtures with more than 50% finer than the 0.075 mm sieve. Coarse-grained
soils are labelled GW, GP, GM. GC, SW, SP, SM, and SC whilst fine-grained soiIs are labelled
ML, CL, OL, MH, CH, and OH.
The name of a fine-grained soil canbc dctcrmincd from the pIasticity chart shown in Figure
1.27 which is divided into two ranges of liquid limit of L for LL < 50% and H for LL > 50%.

Theequation for the A line which separates the clay fraction from the silt and organic matter is
given by:

PI = 0.73(LL - 20) (1.38)

Table 1.6 the Unified Soil ClassificationSystem
shows for engineering purposes according to
ASTM Use of this standard
D-2478. results in a single classificationgroup symbol and group
name except when a soil has 5 to 12% fines or the state of the soil in the plasticity chart fails
into the cross-hatched area. In these cases, a dual symbol is used. e.g. GP-GM, and CL-ML.
Any soil on the borderline of two group symbols is identified by both group symbols separated
by a slash, e.g. SC/CL, CL/CH, and Borderline
GM/SM. symbols are particularly useful when
the LL is close to 50%. These soils can have expansive characteristics and using a borderline

symbol alerts the user of this potential (ASTM D-2487-1998).)))

 Nature of Soils, Plasticim. attd Compaction 35

6O

5O A line\177

\177MH or OH

\1770

ML, \177 \177

0 10 20 30 40 50 60 70 80 90 100 110
Liquid limit (%)

Figure 1.27. Plasticity chart (ASTM D-2487),

Table 1.5.First and second letters of group symbols.

Soil identification First letter of group symbol Second letter of group symbol
Coarse grained soils G: gravel, S: sand W: Well graded, P: Poorly graded
Fine grainedsoils M: silt, C: clay L: Low plasticity, H: High plasticity
Organic soils O L:Low plasticity, H: High plasticity

Highly organic soils Pt No second letter

1.6.3 Description, identification and classification of soils(AS 1726)

The Unified Soil Classification System with minor modifications soils

is used. Coarse-grained
are defined as materials in which more than half of material less than 63 turn is larger than

0.075 mm. The CL group of the original system (LL < 50%)has been divided into two groups
of CL (inorganic clay of low plasticity) with LL < 35% and Cl (inorganic clay of medium plas-
ticity) with 35% < LL < 50%.Field identification procedures are performed on particles finer
than 0.2 nwn after coarse particlesare removed by ]land.

Dry strength (crushing characteristics), A pat of soll is made to the consistency of putty and is
left in the oven or sun or air-dried until completely dry. By crushing it between the fingers, its
strength can then be assessed but only after developing experience from applying the method
to different soil samples with known strength characteristics.

Dilatancy (reaction to shake).A mixture of soil and water (approximately I0 cm 3) is made

with sufficient moisture to make the mixture soft but not sticky, The sample is then placed in)))
36 Soil Mechanics: Basic Concepts and Engineering Applications

Table 1.6. Unified Soil Classification System (ASTM D-2487).

Major division Group symbol Finer\" than 75 grn

GW <5%
Gravelly soils
50% of coarsefrac- GP <5%
tion larger than sieve
GM > t2%
No.4:
2.36mm GC > 12%
Coarse grained
soils
50% > 75 lain SW <5%
Sandy soils
50% of coarsefrac- SP <5%
tion finer
SM > 12%
than sieve No.4:
2.36mm SC > 12%

ML

Siltsand clays CL
LL < 50%
Fine grained OL
soils
50% < 75 gm
MH

Silts and clays CH

LL > 50% OH

Highly organic soils Pt

**For soils with 5%-I2% passing the No. 200 (75 lain) use a dual symbol.

the open palm of one hand and is shaken rapidly, striking vigorously against the other hand.

Water will appear on the surface of the sample, but will disappear if the shaking is stopped and
the sample is squeezedbetween the fingers. The rapidity of appearanceof water during the
shaking process and of its disappearance during squeezing assistin identifying the character of
the fines in the soil. For instance,very fine clean sands give the quickest and most distinct re-
action whereasplastic clay has no reaction. Inorganic silts show a moderately quick reaction.

Toughness (consistencynear plastic titnit). A specimen of soiI approximately 10 cmm vol-

ume is moulded to the consistency of putty and then rolled out by hand on a smooth surface or
between palms into a thread about 3 mm in diameter. The thread is then folded and re-rolled
repeatedly. The sample gradually loses moisture and crumbles when the plastic limit is
reached. The piecesof the crumbIed thread are lumped together with a slight kneading action
which should be continued until the lump crumbles. The tougher the thread near the plastic
limit and the stiffer the lump when it finally crumbles, the more potent is the colloidal clay)))
 Nature of Soils, Plasticity and Compaction 37

Table1.6(continued). Unified Soil Classification System (ASTM D-2487).

Typical names Laboratory classification criteria
Well graded gravels, Co, > 4
gravel-sand mixtures I < Cc< 3
Poorly graded gravels, gravel-sand mixtures Not meeting gradational requirements for GW
Silty gravels, poorly graded gravel-sand-silt Atterberg limits below A- Above A-line
mixtures line or PI < 4 With 4t < PI < 7

Clayey gravels, poorly graded gravel-sand-clay Atterberg limits above A- requires

mixtures line with PI > 7 dual symbols
Well graded sands, gravelly sands Cu > 6
1<Cc<3
Poorly graded sands, gravelly sands Not meeting grad.ational reqmremen!s for SW
Silty sands, poorly graded sand-silt mixtures Atterberg linfits below A- Above A-line
PI < 4
line or with 4< PI <7
Clayey sands, poorly graded sand-clay mixtures Atterberg limits above A- requires
line with PI > 7 dual symbols

Inorganic silts, very fine

sands, silty Or clayey fine Plasticity chart
sands with slight plasticity, rock flour
Inorganic clays of low to medium plasticity, Plasticity chart
gravelly clays, sandy clays, silty clays, lean clays
Organic silts and organic silt-clays of low plasticity Plasticity chart,
organic odour or colour
Inorganic silts, micaceousor diatomaceous fine Plasticity chart
sandy or silty soils, elastic silts, volcanic ash
Inorganic clays of high plasticity Plasticity chart
Organic clays of medium to high plasticity Plasticity chart,
organic odour or colour
Peat and other highly organic soils Fibrous organic matter, will char, bum or glow

fraction in Weakness of the thread

the soil. at the plastic limit and quick loss of coherence of
the lump the plastic limit indicate
below either inorganic clay of low plasticity or organic clay.
Highly organic clays have a very weak and spongy feel at tl\177e plastic limit. Table 1.7 shows a
guide to field identification and the application of dry strelagth, dilatancy and toughness con-
cepts to the group symbols of the Unified Soil Classification System (AS 1726).

Description of a soil. A systematic and standardized order of description similar to Table

1.8(a) (Section A2.3 of AS i726) has been found to be suitable. Most of the descriptive terms
needed in Table 1.8(a) can be found in Tables 1.6 and 1.7. For a description of colourand par-
tlcle shape the following terms are used: the colour of a soil should be described in the moist
condition as (modified if necessary)pale,dark, or mottled and black, white, grey, red, brown,
orange, yellow, green and blue. Equidimensional particles may be described as round, sub-
rounded, sub-angular,or angular as shown in Figure 1.28. The consistency of cohesivesoils
is described in terms of its undrained shear strength (Chapter 4) as shown in Table 1.8(b) (AS
1726) whilst the consistency of non-coheslve soils is describedin terms of the density index
according to Table 1.8(c) (AS 1726).)))
38 Soil Mechanics: Basic Concepts and Engineering Applications

Rounded Sub-rounded Angular Sub-angular

Figure 1.28. Description of panicle shape,

TabIe !.7. Guide to the field identification of coarse and fine-gained soiIs (AS 1726).

Group
Field identification of coarsegrained soils
symbol

GW Wide range in grain size and substantial amounts of alI intermediate sizes, not enough
fines to bind coarse grains, no dry strength.
GP Predominandy one size or range of sizes with some intermediate sizes missing, not

enough fines to bind coarse grains, no dry strength.

GM Dirty materials with excess of non-plastic fines, zero to medium dry strength.
GC Dirty materials with excess of plastic fines, medium to high dry strength.

SW Wide in grain size and

range substantial amounts of aI1 intermediate sizes, not enough
fines to coarse
bind grains, no dry strength.
SP Predominantly one size or range of sizes with some intermediate sizes missing, not
enough fines to bind coarse grains, no dry strength.
SM Dirty materiaIs with excess of non-plastic fines, zero to medium dry strength.
SC Dirty materials with excess of plastic fines, medium to high dry strength.

Field Identification of fine grained soils

Dry strenglh Dilatancy Toughness

ML None to low Quick to sIow None
CL, CI Medium to high None to very slow Medium
OL Low to medium SIow Low

MH Low to medium Slow to none Low to medium

CH High to very high None High
OH Medium to high None to very slow Low to medium

Pt Identified by colour, odour, spongy feel and generally by fibrous texture

Table 1.8(a). Order of description of a soil.

Order of description Details of descri priori

Composition Group symbol, soil name, plasticity or particle characteristics, colour,

secondary components, and minor components.
Conditions Moisture condition (disturbed or undisturbed), consistency (undisturbed).
Structure Zoning, defects, cementing (undisturbed),
Additional observations Soil origin and other matters if significant.)))
 Nature of Soils, Plasticity and Compaction 39

Table 1.8(b).Consistency terms for cohesive soils based on undrained shear strength

Term cu (kPa) Field guide to consistency

Very soft _< 12 Exudes between the fingers when squeezed in hand,
Soft > 12 -< 25 Can be moulded by finger pressure.
light
Firm > 25 _< 50 Can be moulded by strong finger pressure.
Stiff > 50 < 100 Cannot be moulded by fingers; can be indented by thumb,
Very stiff >100 _< 200 Can be indented by thumbnail.
Hard >200 Can be indented with difficulty by thumbnail.

Table 1,8(c), Consistency terms for non-cohesive soils,

Term Density index

Very loose _< 0.15

Loose > 0, 15 -< 0.35
Medium dense > 0.35< 0\17765
Dense > 0,65 _< 0,85
Very dense > 0.85

1.6.4 Summary of British soil classification system for engineering purposes

The soil classification system

British is shown in Table !_9, The qualifying terms are similar
to the Soil Classification
Unified Systeln with more subgroup symbols for both coarse al\177d fine
soils. Poorly graded gravel and sand have been divided into two group symbols by addit\177g Pu
and Pg to represent uniform and gap-graded soils respectively, In coarse soils subgroups have
been definedby adding a third letter (M or C)to well graded or poorly graded gravel and sand,

7O

6O
CV
o\177 50

-\177 4o
CH t
,-\177 30
._o

\177 20

10
CL J MH

\177ML \177 \177

0 10 20 30 40 50 60 70 80 90 100 110 120

Liquid limit (%)

Figure 1.29, Plasticity chart according to British soil classification system.)))

40 Soil Mechanics: Basic Conceptsand Engineering Applications

Table 1.9+ British soil classification system for engineering purposes (BS 5930).

Soil groups Subgroups and Iaboratory identification

GraveI and sand may be defined as sandy gravel Group- Sub- Symbol Fines LL
and gravelly sand, etc. symbol group <6% %

Slightly silty or clayey G GW GW 0-5

gravel GP GPu, GPg
Gravels
more than Silty gravel G-F G-M GWM, 5-
50% of GPM 15

coarse ma- Clayey gravel G-C GWC, GPC

is of
terial
Very silty gravel GF GM GML, etc.
gravel size GC GCL
Very clayey gravel
Coarse (coarser GCI 15-
soils than 2 ram)
35
GCH
lessthan
35% is GCV
finer than
GCE
Slightly silty or clayey S SW SW 0-5
sand SP SPu, SPg
Sands
more than SiIty sand S-F S-M SWM, SPM 5-
50% of Clayey sand S-C SWC, SPC 15
coarse ma- Very silty sand SF SM SML, etc.
terial is of sand SC SCL
Very clayey
sand size SCI 15-
(finer than
SCH 35
2 ram)
SCV

SCE

Gravelly silt t\177G MG MLG, etc.

Gravelly or Gravelly clay CG CLG <35
sandy silts CIG 35-50
and clays 50-70
CHG
Fine soiIs 35% to CVG 70-90
more than 65% fines
CEG >90
35% is
finer than Sandy silt FS MS MLS, etc.
0.06 mm Sandy cIay CS CLS. etc.
SiIt (M-soil) F M ML, etc.
Silts and C CL <3 5
Clay
clays CI 35-50
65% to CH 50-70
100% fines
CV 70-90
CE >90
Organic soils Descriptive letter O suffixed to any group or
subgroup symbol
Peat Pt)))
 Nature of Soits, Plasticity and Compaction 41

Fine soils, and the fine fraction of coarsesoils,are shown by F, and the subgroup symbols of
fines and fine fraction soils are:L (low plasticity),
of coarse I (intermediate plasticity), H (high
plasticity), V (very high plasticity), E (extremely high plasticity) and U (upper plasticity range)
as the second or third letter (Figure 1.29 & Table 1.9).Field identification is based on particle
size,grading, structure, compactness, dry strength, dilatancy, consistency and weathering (BS
5930, 1981).

1.7 COMPACTION

1.7,1 Dry density-moisture content relationship

Compaction is the process of reducing the air content by the application of energy to the moist
soil. Compaction increases the number of particles within a specific volume thereby increasing
the shear strength. Consequently, any displacement due to external loading and surface settle-
ment thereby reduced becauseof the denser structure. Compaction of soil samples in the labo-

ratory is carried out in standard cylinders. Energy is applied by a hammer of standard size and
mass dropping freely from a height on a layer of the sample
standard inside the cylinder. Each
specimen is made in layers with
3 or 5 a specified number of blows depending on the codes
and method of testing. In the field, the energy is applied by means of different types of rollers.
Standard codes specifythis energy per unit volume of soil. The term compactive effort is used
to describe the energy given to a unit volume of soil. Compaction increases the number of par-
ticles per unit volume and dry density is used to indicate the degree of compaction.

Moisture content

Figure 1.30, Soil behaviour under a specified compactive effort.)))

42 Soil Mechanics: Basic Concepts and Engb\177eering Applications

Compactive effort is most effectiveif a uniform mixture of soil and water is used. Dry density
increases pro\177essively with increasing moisture content to a maximum under a constant com-
pactiveeffort. The moisture content correspondin\177to the maximum dry density is called the
optimum moisture content (OMC). Increasing the moisture content beyond this value reduces
the dry density, The common and natural behaviour of a soil-water mixture under a specified
compactive effort is shown in Figure 1.30. The curve representing the experimental resuIts is
termed the dry density-moisture content curve. Maximum dry density and the relevant opti-
mum moisture content are obtained from this plot.
Lambe (1958)explained the shape of the dry density-moisture content curve in terms of the
development and characteristics of the diffuse double layer with increasin\177 moisture content,
At low moisture contents and o1\177the dry side of the optimum moisture content there is insuffi-
cient water to form a double layer and the structure of the fine particles is of a floccuIated
type. The increase in moisture content causes the formation of the double layer and randomly
distributed particles become more orientated, This produces a lubrication effect between the
large particles causing them to slide over each other as the compactive effort is applied, At the
maximum dry density the specimen of compacted soil has a high degree of saturation. Further
increase in the moisture content has a dilution effect and dry density decreases whiIe the de-

gree of saturation remains approximately constant. The effects of compactive effort and mois-
ture content are insignificant on granular soils that have little or no fines. Test points on the
wet side of the OMC may not be established by a standard compactio\177 test due to loss of water
during compaction. A method combing vertical vibration and water gives satisfactory results.
The theoretical relationship between dry density, air content, and moisture content is ex-
pressed by Equation l,. 18, Figure 1.31 shows the plot of this equation for Av = 0% (termed the

\\ Av= 5%

0d max

Moisture conten\177

Figure 1,3I. Theoretical curves and experimental compaction curve.)))

 Nature of Soils. Plasticity and Compaction 43

Moisture content

1,32.
Fi\177m\177re The effect of compactive effort on maximum dry density and optimum moisture content.

zero air curve), 5% and t0% for a certain value of specificgravity of the solids. In practice the

compactive effort cannot de-air the voids completely, thus the dry density4moisture content
curve is always on the left side of the zero air curve. The descending part of the compaction
curve usually has a minimum air content of 3%-5%.If the descending part i\177ltersects the zero
air curve, either the test data or the specific gravity of the solids is wrong. In general the right
side of the zero air curve is an impossible state for experime\177tal data points, It is always con-

venient to plot the air void curves of 5 %, 10%, etc. to understand the development of compac-
tion and the progress of the degree of saturation as moisture content increases. The standard
3
energy applied to soil specimens in the laboratory is 592 kNm per 1 m of soil and was sug-

gested by soil scientist R, R. Proctor during the 1930s. In time, the need for having a denser
and stronger soil was increased an\177 the Proctor method was modified by increasing the com-
pactive effort to 2695 k_Nm per t m\" of soil, Increasing the compactive effort increases the dry
density and reduces the opti\177taum moisture content, as shown in Figure 1.32. Curve C has a

higher compactive effort than curve B. The increase in maximum dry det\177sity is believed to be
as a result of greater orientation of the fine particles under the increased compactiveeffort.
Figure !.33 provides an estimation of the dry densities and optimum moisture contents for the
fine and coarse grained soils defined by the Unified Soil Classification System.

1.7.2 Standard laboratory compaction test

The moulds used for the compaction test, with internal diameters of 101.6 mm or 105 nun and

t52.4 rcml, and the hand-operated rammer are shown in Figure 1.34. Each mould has a remov-

able collar assemblywith a detachable base plate. The mould type is selected depending on the)))
44 Soil Mechanics: Basic Concepts and Engineering Applications

2.3

2.2

2.1
\177=0%

1.6

1.5
1.4
10 20 30
Moisture content (%)

Figure 1.33. Typical compaction curves for coarse and fine grained soils.

size fraction of the particles in accordance to the relevant standard code. In the standard Proc-
tor test compaction of the prepared sampleis conducted in 3 layers with a rammer of 2.5 or 2.7
kg. The number of blows per layer is 25 or 27 for the small size mould and 56 or 60 for the

(c)

Figure ].34. (a) Proctor mould, (b) modified Proctor mould, (c) rammer (Wykeham Farrance).)))
 Nature of Soils, Plasticity and Compaction 45

large This small variation

size mould. in the number of blows and the mass of rammer is due
to minor in the diameter and height
differences of the moulds in the different codes. In the
modified Proctortest the layers are increased to 5 and a heavier rammer of 4.5 kg or 4.9 kg is
used. Mechanicalversions of hand rammers, which comply with most codes, are also avail-
able. Two common types are portable, lightweight vibrating rammer and the stationary auto-

matic compactor. Table 1.10 shows the adopted standards for the compaction test. Details of

the preparation of samples, treatment of oversize material, moisture content intervals, and plot-
ting of the experimental results are in thedesignated codes in the first column of Table 1.10.
The bulk density of the soil in each test is calculated by dividing the mass of the compacted
soil by the volume of the mould. the compacted soil from the mould, a repre-
After removing
sentative sample from the whole height of the specimen is obtained for the determination of its
moisture content. Dry density is calculated from Equation 1.11. The test is repeated 6 times
over a range of moisture contents to establish the dry density-moisture coment curve.

Minimunt and maximum dry density of cohesionlessmaterials. Moisture content has little or
no influence on the compaction characteristics of granular soils (except when the soil is fully
saturated). Their state of compaction can be obtained by relating the dry density to the mini-
mum and maximum dry densities obtained in the laboratory. The test technique requires a vi-
bratory rammer or table and a pouring method. Accordingly the density index (Equation 1.13)
may be calculated in terms of minimum and maximum dry densities if a specified density in-

dex is required. Combining Equations 1.13 and 1.14:

emax -e [3d -Odrnin

ID = _ Odrnza
(1.39)
em.\177 -emin Od Odtrvax -Odrmn

Minimum dry density is obtained using an oven-dried material and is referred to as the dry
placement method. In this method a mould with known volume and a pouring device are se-
lectedaccording to the maximum size of the soil particles, as recommended by the relevant

Table 1.10. Adopted standards for compaction test.

Mould: Diameter (ram), Rammer: No. of layers,
Standard designation
Height (mm\177 Drop (ram), No. of blows per layer
Volume (cm\") Mass (kg)

ASTM D-698 101.6,116.5,944 305, 2.5 3, 25

2100
(PrOctor) 152.4,116.5.
,157, 4.5 5, 25
ASTM D-1557, 101.6, 116.5, 944 305,2.5 5, 25
AASHTO T99, TIS0 152.4,116.5, 2100 457,4.5 5, 56
(modified Proctor)
BS 1377& BS 1924 105.0,115.5.1000 300, 2.5 3, 27
(Proctor)
(modified using AASHTO) 105.0,115.5,1000 450,4.5 5, 27
AS 1289.5.1.1 105.0.115.5,1000 300, 2.7 3, 25
(standard) 152.0. ]32.5, 2,1.00 3, 60

AS 1289.5.2.1 105.0,115.5,1000 450,4.9 5, 25

(modified) 152.0, 132.5, 2400 5, 100)))
46 Soil Mechanics: Basic Concepts and Engineering Applications

code. To prepare a uniform sample with minimum segregation, the mould is filled with mate-

rial, using a funnel or by means of a scoop, in a steady stream, free of vibration and any other
disturbances. After levelling the material the mass of the mould and its contents is measured
for the calculation of dry density. The test is repeatedseveral times until a lowest value for the
mass inside the mould is obtained. Maximum dry density is determined by the wet placement
method using either a vibratory rammer (BS 1377) or a vibratory table (ASTM D-4253, D-
4254, and AS 1289.5.5.1). A diagrammatic section (and view) of a vibrating table permitted
by AS 1289.5.5.1 is shown in Figure 1.35. Generally the apparatus comprises a vibratory table
which oscillates at a nominated rate (say 3600vibrations per minute) and is equipped with two
to five moulds with nominal volumes of 1 litre to 5 litres. A surcharge of 5 to 70 kg is neces-
sary to provide vertical support during vibration. The saturated material is poured into the
mould by means of a funnel or a scoop until it is filled and overflowing. If necessary, water is
added to ensure full saturation. The mould is vibrated during the filling process with smaller

amplitude of vibration. The surcharge is loweredto the surface of the sample and the vibrator
is set to a required amplitude for approximately 10 minutes or until the settlement of the sur-
face has stopped. After removing the surcharge and levelling off the material, the contents of
the mould is placed in an oven to dry to a constant mass. The maximum dry density is then
calculated by dividing the dry mass by the volume of the mould.

1.7.3 Compaction in the yTeld

Compaction in the field may be divided into the two categories of deep and surface compac-
tion depending on the thickness of the compacted layer. Deep compactionis basedon vibro-
compaction techniques, dynamic compaction and the use of explosives. A detailed study of
deep compaction is beyond the scope of this book, however, due to its increasing applications
in surface and near-surface compaction, a brief description of dynatnic compaction is included
at the end of this section.

Surcharge

_ r\177

\177.... !

''\177''\177\177
\"'\177\"\177\"\177\"\177C o n cre t e floor

Figure 1.35. Apparatus for vibratory compaction.)))

 Nature of Soils, Plasticity at\177d Compactiott 47

Surface colnpaction includes tile compaction of layers in 0.15 m to 0.3 tll laym s. however, by

t\177sing vibratory techniques al\177d dynamic compaction, the depth of the compacted layer may be
increased to 1.0 m or more in granular soils\177

The maximum dry density obtained i\177laboratory condition,s \177sachievable in the field by us-

ing diftErent types of compactors Ihat employ diffelet\177t concepts in the applicalion of the com-

pactive effi)\177. Degree of co\177npaction is a means of comparing the fie{d de\177sity with laboratory
results and is defined us the ratio of the dry densily in the field to the lnaximum dry denstty
obtained in lhe laboratory. I1\177most co\177struct\177o\177 works Ibc degree of compactiol\177 is specified as
()5% or ii/o1c,ilSifally without any designation of tile method of compaction. The hulk and dty

density in che field are measured using one of the methods described in Section 1.3. If the
compacted sozl does not expand or cunsolidate altm the field compacuon is c\177m\177pleted, its dry
density \177emai\177s constanl reg\177u'dless of tl\177e magnitude of its moisture contenl. In the field, the
moisture co\177ltent lnay vary from zero to a iI\177ax\177n\177umvalue which is the intersection of a hori
zontal line drawn from the \177i\177aximum dry density on tile expevi\177nentaI curve with tim theoreti-
cal zeroair curve. Stability, sludies of compacted soils show tba\177 a COlnpacled fill may tttidcrgo
compression or consolidation due to self weight resulti\177g a substantial surface settlemm\177t or
failure (Lawton eta{., 1989: [\177randon ct hi., \177990). In order to predict the stability of a corn
patted fill i\177is necessary to perft\177rm lah(ntitory slrength and c\302\242msolidat\177m tests <m the

pacted speci mens.

Compactionequipme\177t ct)llllllollly used i\177the field are:

\177
Sheep'\177v.{bot roller. This \177otler consists drum with studs or feet 40 to 100 cm m area
of a steel
projecting some 200 to 250 mm It ts either pushed or pulled by a t\177actor or is se{f-plopetled
unit. A sheep's fo\177)t roller is ctmvcnicnt for col\177es\177ve soils and applies compactum by a cotl)bt-
nal\177on of tamping aod kneadi\177g. Contact applied
i)\177es>t\177xe\177 by the projecled studs are high arid

vary from 700 to 4000 kPa_ The thicMzess of a c\177m\177pac{ed layer is of the t>rder of 0.15 m to 0\1773
m. Bo\177d between Iwt) compacted layers is very smmg due action
\177o\177l\177e of the studs. The type
of the roller (litter or heavy) a\177d Ihe number o\177passes must be determit\177cd by the site engineer
attin conducting a sui{ahle number of field density tests.

Pt\177cut\177tatic This
rol[e\177 \177.
t\177'\177e\" type of lolIer is either towed o\177is a self-propelled unit cOn\177lno\177lly

equipped with an outside f\177Ited for weight

box adjt\177stn\177enl. It usually has IWO axles with 2 to 5
(or morej tyrcs per axle and \177hc hack axle is designed to overlap \177e area compacted by the
['ro:\177t axle. are sometimes eqmpped with a wobb]i\177g mechanism to cover the lower
The wheels
surfaces. The tyres have Ilat treads a\177d theretBre cam\177{}t apply sideways displacements Io the
soil particle\177_ This con\177pactor is suitable fo\177 mauy differem soils and applies compaction by
k::ead\177ng onl>. Compacted layers vary flom 0.15 m to 0.5 m dcpc\177ding on the mass of the
ct)mp\177ct(\177r. Usually only a few passes are necessaryIo achieve maximum d\177*ydensiu.

Vibratory rollers. These are applicableto many types of \177,OllS but are primarily designed for

granular matermls. 1\177tim sI]lUOlI\177, drum wbrators the vibrati\177g mecht, co\177isists
ms\177I\177 of ecceutri
call 5' ammged weigh,s (Terzaghi ;rod Peck, 1%7) which in rotation apply a vertical vibr\177,tion

the drum.
t\177.\177 In pneumatic-tyred vibratory compactors lhe viht'ating mechanism is applied to)))
48 Soil Mechanics: Basic Concepts and Engineering Applications

the axle of the vibrating plate compactors a plate or platesthat are in contact with
unit. In the
the ground surface are
subjected to vibration: however the depth of influence is not as great as
smooth drum and pneumatic tyred vibrators. These \177atter two machines compact the soil to a
depth of up to 1.0 m, but an effective depth of 0.3 m is generally accepted if the soil consists
of sir and clay. Vibratory rollers are available in towed or self-propelled units,

Dynamic compaction, In a weight of 100 to 200 kN is dropped from a height

this method of 15

to 40 m using Higher weights of 500 kN may also be used if a deeper compac-

crawler cranes.
tion is required. Establishing the number of passes and drops per pass needsa careful in-situ
study. Experience has shown that granular materials require more energy than fined grained
soils. In the latter the applied energy increases the pore water pressure and adequate time is
needed for dissipation of any excess pore pressure. In the laboratory dynamic compaction is
modelled using the centrifu_\177e method in which the force can be increased up to 100g (g = 9.81
rrdsec-).Here dynamic compaction is provided by dropping a steel rod several mm in diameter
and approximately 100 mm in length on the surface of the sample. The centrifuge machine for
this purpose should be capable of acceptinga 0.5 m cube of soil.

1.7.4 Application of compaction test in sol1 stabilization techniques

The standard compaction test is also used in soil stabilizatioq to investigate the effect of the

/ Zero air voids (soil)

1.9
rUntreatedso,i
x\177/

1.75

1.7

1.65 Field

| 4% Quicklime
1,6 lab mix

[
1,55 t , , , , ,

8 10 12 I4 16 18 20 22 24

Moisture content (%)

Fi\177v.tre 1.36, Compaction curves for a clay soil stabilized by lime (Aysen et al.. 1996).)))
 Nature of Soils, Plasticity and Compaction 49

stabilizer on the maximum dry density and optimum moisture content. Depending on the type
of the material and the stabilizer (e.g. lime or cement) the maximum dry density may increase
or decrease s\177ightly or moderately causing variation in the optimum moisture content in the
reverse direction. In these cases the strength of the treated material is not represented by the
maximum dry density and the increase ila strength is mostly due to the formation of the bonds
between the particles. Investigating the optimum moisture content is vital to ensure that suffi-
cient water is available for hydration. An example of lime stabilization of clay material is
shown in Figure 1.37 (Aysen et aL 1996).

Example 1.9
The following results were obtained from a standard compaction test. Determine the optimum
moisture content and maximum dry density. Plot the curves of 0%, 5% and 10% air content
and give the value of air content at the maximum dry density. The volume of the standard
3
mould is 1000cm and Gs = 2.7.

Mass (g) 1768 1929 2074 2178 2\17706 2052 2007

w (%) 4 6 8 10 12 14 16
Solution:

Calculate dry density for each test and tabulate the results. Sample calculation for w = 4%:
0 =M IV = i768/1000 = 1.768 Mg/m3, 13 = 91(1 + w) = 1.768/(I + 0.04)= 1.70Mg/m 3.
d

9d (Mg/m3) 1.70 i.82 1.92 i.98 1.88 1.80 1.73

w (%) 4 6 8 10 12 14 16

To plot the constant air content curves use Equation 1.18 and tabulate the results. The results
3
are shown in Figure 1.37. It is seen that the maximum dry density is 1.98 Mg/m and optimum
moisture content is 10%.

w (%) 4 6 8 10 12 14 16 18 20
Pe
3
A = 0% 2.44 2.32 2.22 2.13 2.04 |.96 1.88 i.81 1.75
(Mg/m) A = 5% 2.32 2.20 2.1i 2.02 1.94 i.86 1.79 \177.72 1.66
A =10% 2.20 2+09 2.00 1.92 1.84 1.76 1.69 1.63 1.57

Calculate the air content at maximum dry density using Equation 1.18:
9a = GsOw WGs) = 1.98,2.7\303\227
= 6.86%.
- A = 0.0686
(1- A v)/(1 + 1.0(1 v)/(1 + 0.1\303\227
2.7) --\177A v

Example 1.I0

For a soil with Gs = 2.65, the following results were obtained from a standard compaction test:

Sample number [ 2 3 4 5
Moisture content
U\177)
16.2 16.7 19.0 20.4 21.6
Dry densily (Mg/m-) 1.580 1.620 t.647 1.605 1.566)))
50 Soil Mechanfcs: Basic Concepts and Engineering Applications

2,5

2,4 \177 =5%

2.3 L
2.2
2.1
1.98

\177.8

1.7

1.6
\177
10%
1.5
2 4 6 8 10 12 14 16 182022
Moisture content (2/\302\260)

Figure 1.37. Example 1.9.

To obtain two points close to the maximum

extra dry density two standard compaction tests
were performed and Determine
tabulated.
(a) Pdmax and
Woet and plot the zero air curve, (b) a
0.3 rrt layer (\177fthis soil is compacted in the field to its maximum dry density. After some time
the moisture content of the layer is measured to be 16%. How much water in terms of
n\177atural
m/m- is needed to make the layer fully saturated?

Sample number 6 7
Mass of mould 4- compacted wet soil (kg) 8.966 8.974
Mass of mould (kg) . 7.0 7.0
Volume of mould (cm \177) 1000 1000
Mass of sub-sample taken from mould (g) I78.8 155,8
Mass of sub-sample after drying (g) 152.3 131.8
Solution:

(a) For sample number 6:

9 = M/V = (&966-7.0)/1000 = 1.966
Mg/m 3, w = (178.8 -152.3)/152.3 = 0,174,

Od = 9/(1+ w) = 1.966/(1 + 0.174) = 1.675Mg/m 3.

For A\177,= 0% use Equation 1.17; results are tabulated below and are shown in Figure 1,38 from
3
which P\177'max
= 1.68
Mg/m and wop\177 = 17.7%.
(b) =
Gs p + wG = 1.68, 2,65x 1.041+ 2.65 w) = L68 --\177 w = 21.8 %.
p,\177 \177./(1 s)
3
Aw = 21.8-16.0 = 5.8%, AM\177,.
= 1.68x0.058 = 0.097 M\177m or AV w =0.097 \177n3/m 3.

AVw( 0.3
= 0.029m3/m
m layer) = 0.097:<0.3 2, h\177.
= 0.029/1.0 = 0.029 trdm 2.)))
 Nature of Soils, Plasticity and Compaction 51

1.9

1.8

1.7

1.6

= 1 7.7%
Wop t
i
! [ I I

15 17 19 21 23
Moisture content (%)

Figure i.38. Example 1.10.

Test W (Mg/m3)
9,\177 Test W Pe (Mg/m3)
points (%) A = 0% Test points (%) A = 0% Test
1 16.2 1.854 1.580 3 19,0 1.762 1.647
2 16.7 i.837 i.620 4 20.4 1.720 i,605
6 17.4 1.814 1.675 5 21\1776 1.685 1.566
7 18.2 1.788 1.670

1.8 PROB LEMS

1.1 The following data are given for a specimen of clay soil: M = 221
g, Ms = 128 g, Gs=
2.7, Sr
= 75%. Determine the total volume and the porosity of the specimen.

Answers: 171,4 cm 3, 72.3%

1.2 Dry soil (Gs = 2.71) is mixed with 16% by weight of water and compacted to produce a
cylindrical sample of 38 mm diameter and 76 mm long with 6% air content. Calculate

the mass of the mixed soil that will be required and its void ratio.
Answers: 177.7 g, 0.525
1.3 During a field density test 1850 g of soil was excavated from a hole having a volume of
900 cm3. The oven-dried mass of the soil was 1630 g. Determinethe moisture content,
dry density, void ratio and degree of saturation. Gs = 2,71.
Answers: 13.5%, 1.81 Mg/m 3. 0.496, 73.8%)))
52 Soil Mechanics: Basic Conceptsand Engineering Applications

1.4 A soil specimen has a moisture content of 21.4%, void ratio of 0.72, and G.\177
= 2.7. De-
termine: (a) bulk density and degree of saturation, (b) the new bulk density and void ratio
if the specimen is compressedundrained until full saturation is obtained.

Answers: 1.905 Mg/m 3, 80. i%, 2.08Mg/m 3, 0.578

1.5 The moisture content of a specimen of a clay soil is 22.4%.The specific gravity of the
solids 2.7 (a)
is t. Plot the variation of void ratio with degree of saturation and calculate
the void ratio, and the dry and wet densities at 50% saturation, (b) a sample of this soil
with initial degree of saturation of 50% is isotropically compressed to achieve a void ra-
tio of 0.55. Calculate the volume change in terms of percentage of the initial volume.
How much of this volume change is due to the outward flow of water from the sample?
Answers: 1.214, 1.224Mg/m 3, !,50 Mg/m 3, 30.0%, 2.6%
1.6 The results of a particle size analysis are:

Sieve size (ram) Mass retained (g) Sieve size (mm) Mass retained (g)
63 0.0 4+75 50
37.5 26 2.36 137
19,0 28 1.18 46
13.2 18 0.6 3!
9.5 20 0.212 34
6,7 49 0.075 30

The total mass was 469 g. Plot the particle size distributioncurve and determine the co-
efficient of uniformity, coefficient of curvature and soil description.
Answers: 13.3, 3.2, GW
1.7 The following data were recorded in a liquid limit test using the Casagrande apparatus.
Determine the liquid limit of the soil. Classifythe soil assuming PL = 19,8%.

Number of Mass of can (g) Mass of wet Mass of

blows soil + can (g) soil + can (g)
8 11.80 36.05 29.18
16 13.20 34.15 28.60
27 14.10 36,95 31.16
40 12.09 33.29 28.11

Answers: 34.2%, CL-ML

1,8 The recorded data in a liquid limit test using the cone penetration method are as fotlows.
Determine the liquid limit of the soil.

Conepenetration (mm) 14.1 18.3 22.1 27.2

Moisture content (%) 28.3 42.2 52.4 63,4
Answer: 47.0%)))
 Nature of Soits, Plasticity and Compaction 53

1.9 The maximum and minimum void ratios for a sand are 0.805and 0.501 respectively. The
field density test performed on the same soil has given the following results: p = 1.81
3
Mg/m , w = 12.7%. Assuming Gs
= 2.65,
compute the density index.

Answer: 0.50
1.10 The following results are \1773btained from a standard compactio\177 test:

Mass of compacted soii (g) 1920.5 2051.5 2138.5 2147.0 2120\1770 2081.5
Moisture content (%) 11.0 12.1 12.8 13.6 14.6 16.3

The specific gravity is 2.68, and the volume of the compaction mould
\1773fthe solids is
3
1000cm . Plot the
compaction and obtain the curve
maximum dry density and \1773ptimum
m\1773isture content. Plot also the 0%, 5% and 10% air void curves. At the maximum dry
density, calculate the void ratio, degree of saturation and air content, If the natural mois-
ture content in the field is 11.8%, what will be the possible maximum dry density if the
soil is compacted with its natural moisture content?
3
Answers: 1.907 Mg/m 3, 13.0%, 0.407, 85.6%, 4.1%, 1.80Mg/m

1.9 REFERENCES

ASCE, 1962. Symposium on grouting, Journal SMFE, ASCE 87(SM2):1-145.

ASTM D-2487. 1998. Standard classification of soils for engineering purposes (Unified Soil Classifica-
tion System). American society for testing and materials. West Conshohocken, PA.
Atkinson, J. 1993. An introduction to the mechanics of soils and foundations. London: McGraw-Hill.
rd . . .
Australian Standard.1993. Geotechnical site investigations. 3 edmon. Austraha NSW: Standard Asso-
ciation of Australia.
Aysen, A.. Ayers, R., & Ricks, J. 1996.Soil stabilisation with cement and atuminiurn hydroxide. Re-
search Report No CE96/5, ISBN 0909756295. Toowoomba:The University of Southern Queensland.
Bohen, H.L.,McNeal, B.L, & O'Connor, G.A. 1985.Soilchemistry, New York: John Wiley+
Bowles, I.E. 1996, Foltndation analysis and design. New York: McGraw-Hill.
Brandon, T.L, Duncan. J.M. & Gardener. W,S. 1990.Hydrocompression settlement of deep fills. Jour-
nal SMFED, ASCE 116(GT10):1536-1548.
Brewer, R. 1964. Fabric and mineral analysis of soils. New York: John Wiley & Sons.
Brewer, R. & Sleeman, J.R. 1988.Soilstructure and fabric. Australia, Adelaide: CSIRODiv, Soils.
British Standard 5930. 1981.Codeof practice for site investigation. London: British Standards Institu-
tion.
Casagrande, A. 1948. Classifications th and identifications of soils.Translated: ASCE 113: 901-991,
Craig, R.F. 1997.Soil mechanics. 6 edition. London: E & FN SPON.
FitzPatrick, E.A. 1983. Soils, their formation, classifications and distributions. London: Longman.
Fookes,P.G.& Vaughan, P.R. 1986. A handbook of engineering geomorphology. Surry: Surry Univer-
sity Press.
Isbell, R.F. 1996. The Australian soil classification. Australia, Collingvvood:CSIROPublishing.
Jenny, H. 1941. Factors of soilformation. New York: McGraw-Hill.)))
54 Soil Mechanics: Basic Concepts and Engineering Applications

Larnbe, T.W. 1958. The engineering behavior of compacted clay. Journal SMFE, ASCE 84(SM2), May
1958.
Lawton, E,C., Fragaszy, R.J. & Hardcastle J.H. 1990. Collapse of compacted clayey soils, Journal
SMFED, ASCE 115(GT9): 1252-1267.
McLaren, R.G. & Cameron. K.C. 1996.iron exchange in soils and soil acidity. Soil science: sustainable
production and environmental protection. Auckland: Oxford Press.
Mo\177re, C.A. 1970a+ Suggested method for application of X-ray diffraction of clay structural analysis to
the understanding of the engineering behavior of soils. Special procedures for testing soil and rock

for engineering purposes. Philadelphia: ASTM publication.

Moore.C.A. 1970b. Suggested techniques for measuring the fabric of engineering soils+ Special proce-
dures for testing soil and rock for engineering purposes. Philadelphia: ASTM publication.
Oilier, C.D. 1975.Weathering, London: Longman Group Ltd..
Rowe, P.W. 1962. The swess-dilatancyrelation for static equilibrium of an assembly of particles in con-
tact. Proc. roy. soc. A, 269: 500-527.
Sherard, J.L. 1953.The influence of soil properties and construction methods on the performance of ho-

mogeneous earth dams. US bureau of reclamation. Tech. Memo (645).

Skempton, A.W. & Northy, R.D. 1953.The sensitivity of clays. Geotechnique 3(1): 30-35.

Sposito,G. 1989. The chemistry of soils. Oxford: Oxford University Press.

Tan. K.H. 1994.Environmental soil science. New York: Marcel Dekker inc..
TerT.aghi. K. & Peck, R. B. 1967. Soil mechanics in engineering practice. New York: John Wiley.
White, R.E. 1979,h\177troduction to the principles and practice of soil science.Melbourne: B lackwell Sci-
entific.

Whittig, L.D. 1964. X-ray diffraction techniques for mineral identification and mineralogical composi-
tion, In, Methods of soil analysis, American society of agronomy monograph. Part 1, Chapter 49.
Whyte, I.L. 1982.Soil plasticity and strength: a new approach using extrusion. Ground engineering
15(i): 16-24+
Yong,R.N, & Warkentin, B.P. 1966+ Introduction to soil behaviour. New York: The Macmillan Com-

pany.)))
CHAPTER 2

Effective Stress and Pore Pressure in Saturated Soils

2.1 INTRODUCTION

In a saturated soil with no water equilibrium exists betweenstressesin the

movement,
solid and liquid phases. The stressrelated
the internal forces acting
to on the contact

points of the particles is termed the effective stress whilst the stress within the liquid phase
or water is termed pore pressure.The combination of these two stresses representsthe to-
tal stress at a point. When a saturated soil is subjected to external forces, the state of equi-
librium is altered and changes the effective stresses and pore pressures from their initial
values to new values. This increasein the pore pressure, called the excess pore pressure,
dissipates in time depending on the drainage conditionsand permeability of the soil. This
pressure dissipation, in turn, causes a reduction in the volume of the soil, which is pre-

dicted by a mathematical-physical model called the consolidation analogy.

2.2 STATE OF STRESSAT A POINT DUE TO SELF-WEIGHT

2.2.1 The effective stress concept

Terzaghi's concept of effective stress states that:

= \302\242\177'+
\302\242\177 u (2.1)

where \177 is the total normal stress at a point in a specified direction or plane, o-' is the ef-
fective normal stress on that plane resisted by the particles and u is the pore pressure act-
ing on the plane. To fully understand the concept of effective stress, considera cylinder

containing coarse granular material in a dry or a wet condition subjected to a force N\"
(Figure 2. l(a)). Assuming there is no friction between the soil particles and the inner sur-
face of the cylinder, the entire force N' is transmitted completely to the base. The average
vertical stress at the base of the cylinder is the ratio of N' to the internal cross-sectional
area of the cylinder. This stress is much less than the stresses created at the contact points.
As there is no pore pressure in the system the total stress and effective stress at the base of
the cylinder are identical. Consider, now, that the cylinder is filled with water until the soil
is fully saturated and that a force U is applied through another piston as shown in Figure
2.1(b). The loading plate that carries force N' has a hole h for the passage of water to en-

sure a uniform pore pressurethrough the whole system. The fraction of the base area oc-

cupied by the contact points between the particles and the base may be ignored without
55)))
56 Soil Mechanics: Basic Concepts and Engineering Applications

SOIL
' SOIL

(a) (b)

Figure 2.1. Illustration of the concept of effective stress.

significant error. If the self-weight of the components are negligible (in this example) then
the total vertical stress at the bottom of the cylinder is the sum of the pore pressure caused
by U and the effective vertical stresscausedby N'. In this example the loading systems for
the particles and the water are independent with no apparent relationship. However, in a

real soil, N' and U may have the same source,changing the values of effective stresses and
pore pressures as time related functions. Equation 2.1 satisfies the states of stressfrom the

initial state through to the final state. Figure 2.2 shows the average sectiona\177 area ab that is
equivalent to the area of the base of cylinder in the example above. The effective normal

stress on the plane ab is equal to the sum of the components of the forces FI to F7 perpen-
dicular to the plane ab divided by the sectional area represented by ab,

2.2.2 Effective stress and pore pressurewithin unconfined and confined aquifers
A schematic illustration of the water conditions is shown
different types of ground
in Fig-

ure 2.3 (Bell, 1993).In an unconfined of the water, known

ground water as the
the surface

phreatic surface, water level or water table, is in equilibrium with the atmospheric pres-
sure; that is, the aquifer is not overlain by a material of lower permeability. In a confined

aquifer both the upper and lower boundaries are confinedwith impermeable strata, and the
water have sufficient
may pressure to rise above the overlaying stratum (artesian condi-
tions). In a leaky aquifer, a layer of low permeability material separates the confined and

unconfined aquifers. This layer slowly transmits the water and creates a water table in the
unconfined region and a piezometric level (the level of water rising in a piezometric

standpipe) in the confined region.

i FI F2 ,-

Figure 2.2. The concept of effective stress.)))

 Effective Stress and Pore Pressure in Saturated Soils 57

Ground Surface Ground

\177 Surface

Water table : Water table

Unconfined aquifer ; i
Unconfined aquifer
\177\177\"\177/////////.,\177 !. , ,

Impermeable Stratum \177/\177f\177

! \" \" '
. Aquitard. \342\200\242
.. , \"\",. ,. \"
\177.,./.,.\177///.,.//\177,z,,.,./,.,./,z*.//,,\"//////////\177 \177 \177N

Piezometric level
! .\177
Confined aquifer ! \177 Leaky aquifer

\177\177\177\"\177/\177/////'/\177. !\177......,///////////.,\177
\177'\177
i.\177um i\177 Impermeable Stratum \177

:
(a) (b)
Figure 2.3. (a) Unconfined and confined aquifers, (b) leaky aquifer.

Assuming that there is no movement of water within the voids of the soil, the effective
vertical stress \177'v at depth z for three possible positionsof the water table (Figure 2.4) is
given by:

(\177v=Cfv --U, =\1773gz--u

\302\243Yv =TsatZ--U
If the water table is at the ground surface (Figure 2.4, case a), then

= z - -
z = (\302\245.\177a\177 =
\177'v Y sat yw Yw)z \302\245'z (2.2)

In case b the water table is below the ground surface and therefore
- -
Zw) + \302\245'zw
= + =
\177'v Ysa\177Zw Yary (Z Zw) -YwZw \302\245&y(z
(2.3)

In case c where the water table is above the ground surface:

(\177v ('\177sat z
+ Ywhw) -\"/w (z + hw) = \177
= tz

which is identicaI to Equation 2.2 and shows that o-'v at a point below the water table is
constant regardless of the position of the water table above the ground surface.
For a soil section composedof n layers each having a thickness of hi:

cy v =
Yyz -u = eihi (2.4)
i=1

WT (case c)

__ a)
\177se \177

WT (case b)

Figure 2.4. Effective stress in a saturated soil.)))

58 Soil Mechanics: Basic Concepts and Engineering Applications

where Yei is the effective unit weight of each layer. In dry and partially saturated layers
is the dry or wet unit weight; in saturated layers with hydrostatic water pressure, Yei is the

submerged unit weight. In the presence forces due to

of seepage movement of the water
through the voids the hydrostatic based equations become invalid and the effective stress
must be evaluated using the main definition given in Equation 2.1.

Example 2.1
A soil section is comprised of two layers with the following properties: Soil 1 (0 to 3 m):
= 2.0 Mg/m3; 9 3
Pdry
= 1,8
Mg/m 3, 9sar soil 2 (3 m to 8 m): Psat = _.1
Mg/m . The water ta-

ble is t.5 m below the ground surface. Plot the total vertical stress, pore pressure, and

effective vertical stress over the soil section.

Solution:
On the ground surface, ov = 0.0, u = 0.0, CCv = 0,0.
At z = 1.5 m: c\177 v
= 1.8\303\2279.81\303\2271.5 = 26.5 kPa, u = 0.0, \177v = 26.5 kPa.
At z = 3.0 m: c\177, = 1,8\303\2279.81\303\2271.5 + 2.0\303\2279.81\303\227i.5 = 55.9 kPa.
u = 1.0\303\2279.81\303\2271.5 = 14.7 kPa, \177v = 55.9 -14.7 = 4\177.2 kPa.
At z = 8.0 m: c\177 v
= 2,1\303\2279.81\303\2275.0 +55.9 = 158.9 kPa, u = 1.0\303\2279.81\303\2276.5 = 63.8 kPa,

c\177,
= 158.9 -63.8 = 95. i kPa. The results are presented in Figure 2.5.

Example 2.2

A soil profile is shown in Figure 2,6(a). Plot the distribution of total vertical stress, pore
pressure, and effective vertical stress up to a depth of 12 m,

Solution:
Calculate and saturated densities in soil 1 : Vv + = 1, e = V / V Thus:
the dry V
s v s.
\177
=
l/(l+e) and V
v =e/(l+e), for e = 0.6: Vs = 0.625 m 3, Vv = 0.375 m 3.

Stress (kPa)

0 50 1 O0 150 200
0
I
2 oil \177

g4

6
7
8 \"Pore pressure Total
Effective
9

Figure 2.5. Example 2.1.)))

 Effective Stress and Pore Pressure in Saturated Soils 59

Stress (kPa)
0 50 100 150 200 250 300
0.0m \177 0
1
WT
\177.0 m 2
3 Soil 1
Soil 1 Gs = 2.65, e = 0.6
4
\177
5.0 m 5

Soil 2 = 2.13 3 \177

6
Psat Mg/m Soil 2
\177. 7

8.0 m
3 9
Soil 3 Psat
= 2.18
Mg/m
10

re
i?erses\177\177u \177oi,3
12.0 m 12 \\\177
Effecli,\177e Tolal
(a)

(b)

Figure 2.6. Example 2.2.

Pdry
= Gsx pw x Vs= 2.65 x 1.0 x 0.625= 1.656 Mg/m 3,
3 3
M = 1.656 Mg + 0.375m x 1.0 Mg/m = 2.03 ! Mg,
= 2.031 Mg/m 3.
Psat
The results are tabulated below and presented in Figure 2.6(b).

Depth (m) o\177,(kPa) u (kPa) o\"v (kPa)

0 0.0 0.0 0,0

2 32.5 0.0 32.5

5 92.3 29.4 62.9

8 155.0 58.9 96.1
12 240.5 98.1 142,4

Example 2.3

Resolve Example2.2 assuming that: (a) the water table is at the ground surface, (b) the

water table is 2 m abovethe ground surface.

Solution:
Calculations are summarized in the table below and are presented in Figure 2.7.

(a) (b)
Depth (in)
cr v (kPa) u (kPa) o% (kPa) \177v (kPa) ,\177
(kPa) o\"v (kPa)

0 0.0 0.0 0.0 19.6 19.6 0.0

5 99.6 49.0 50.6 119.3 68.7 50+6
8 162.3 78.5 83.8 181.9 98+ i 83.8
12 247.8 117.7 130.1 267.4 137.3 13OI

Note: effective stresses at both cases are identical.)))

60 Soil Mechanics.\" Basic Concepts and Engineering Applications

Stress (kPa)
0 50 100 150 200 250 300

Figure 2.7. Example 2.3.

Example 2.4
The lower sand layer in the soil profile of Figure 2.8(a) is in an artesian condition. Calcu-
late (a) \177\"vat the top and base of the clay layer, (b) the height of water in the standpipe for
c/v = 0 at the base of the clay layer, (c) the maximum depth of the proposed excavation
(Figure 2.8(b)).

Solution:

(a) At z = 4.0 m: c\177 v

=
(1.7x2.0+ 2.1x2.0)\303\2279.81
= 74.6 kPa,
u = 1.0\303\2279.81\303\2272.0 = 19.6 kPa, cr\177,
= 74.6-19.6 =55.0 kPa.

Standpipe
H

3
WT Sand Pan/= 1.70 Mg/m
2.Ore \177
= 2.10 3
Mg/m
4.Ore \177 Osat

= 2.10 3
Clay Psat Mg/m Clay

8.0 m \177
Sand Sand
(b))))

Figure 2.8\177 Exarnp]e 2.4.

 Effective Stress and Pore Pressure in Saturated Soils 61

At z = 8.0 m:o'v + 2.1x4.0)x9.81= 157.0

= (1.7x2.0 + 2.1x2.0 kPa,

u = 1.0xg.81xl 1.0 = 107.9kPa, = 49.1 kPa.

\177
= 157.0- 107.9
(b) At z = 8.0 m cCv= c\177v- u = 0, \177 = u = 157.0
kPa, u = 0,\177xg \303\227h = 157.0 kPa.

h = 157.0 / 9,81 = 16.0m. Height above the ground surface = 16.0- 8.0= 8,0m.
(c) With the arrangement of Figure 2.8(b): o\"v = c\177v- u = 0, c\177v= u, assume z > 2.0 m,
\177v
=
[2.1(4.0
- z) + 2.1x
4]x9.81= 1.0 x i 1.0 x9.81 = 107.9kPa, z = 2.76 m > 2.0 m.

2.2,3 Negative pore pressure due to capillary rise

Capillary rise results from the combined actions of surface tension and intermolecuIar
forces between the liquid and solids. Noting that the pressure on the water table level is
zero, any water above this level must have a negative pressure. In soils a negative pore
pressure increasesthe effective stresses and varies with the
degree of saturation and by
drying or wetting processes. The phenomenon can be demonstrated by immersing a glass
tube of small inner diameter into water as shown in Figure 2.9. The capillary water has a

hydrauIic continuity table and rises to a certain

with the water height where equilibrium
between the weight of the water column and the vertical resuItant of the surface tension
forces is achieved.The height of the capillary column hc is definedby:
h = 4T d) (2.5)
c cos0/(\302\245w

where T is surface tension

the (0.074 N/m at 20\302\260C), 0 is the angle of surfacetension with

the d is the diameter

vertical, of the tube, and Yw is the unit weight of water. The water

surface in the tube (meniscus) is assumedto be a portion of a sphere. For pure water and a
glass tube of very smalI diameter, 0 = 0, the meniscus is hemispherical and:
=
h
c 4T/(yw d) (2.6)
The maximum negative pore pressure is:
u = = 4T/d
c hey w (2.7)

Any of Equations
application 2.6 and 2.7 to a soil raises questions concerningthe tube
concept, the definition of d and the limitations of the equation. The capillary process starts
as water evaporates from the surface of the soil. The capilIary zone is comprised of a fully

WT

Figure 2.9. Cylindrical capillary robe.)))

62 Soil Mechanics.\" Basic Concepts and Engineering Applications

saturated layer with a height of usually less than hc, and a partially saturated layer overlain
by wet or dry soil. In the partially saturated layer the water and air within the voids are
both continuous. Terzaghi & Peck (1967)suggested that the pore pressure in this layer
should be treated as a negative hydrostatic pressure with its maximum expressedby Equa-
tion 2,7. Negative pore pressureresults in an increase in the effective stress and is termed
soil suction.
The boundary between the capillary and the gravitational water is ill defined and may
not be determined accurately(Bell,1993). In the field, a borehole that can create a well is
a simple way of determining the water level, but suction measuring devices are also used.
The diameter of the equivalent tube d l\177rgely depends on the particle size distribution of

the soil and the assumed value of d = eDlo has been generally accepted by soil scientists.
Estimation for the range of capillary rise can be made by assuming 0.2D10 < d < DI0
(Powrie, 1997).
In granular materials (gravels and sands) the amount of capillary rise is negligible while

in silty soils the water may rise up to several metres.The rise and fall of the water table

and gravitational drainage from the ground surface can affect the height of capillary rise
and may create different types of unsaturated zones above or close to the capillary zone.

Immediately above the level he, the suction in the water exceeds the tensile strength of the
water and generates vacuum conditions and vapour pressure (Klausner, 1991). As the ele-
vation increases beyond the capillary rise, suction increases and the soil moisture and de-
gree saturation
of decrease. Eventually continuous air voids with atmospheric pressure
evolve with typical behaviour shown in Figure 2.10. Locally created capillary menisci be-
tween the small particles causes an increase in strength due to the increase in effective
stress. Simple examplesinclude the stability of sand castles made by children at the beach
or the ability of a moist sand layer to resist shear stresses induced by vehicular tyres. The
apparent increase in the strength of a fine granular material disappears with full saturation
or full drying. In the latter instance, the effective stress increases due to the reduction in
the radius of the meniscus and the soil shrinks. The process continues until the shrinkage
limit is reached and soil volume becomes constant. The strength increases as the drying

process continues and reaches its maximum in full dry conditions. This behaviour is re-
versible in fined-grained soils (Section 1.5), that is, the soil volume increases (after the

shrinkage limit), decreasing the strength of the soil.

Ground surface

(-) Pore pressure

Capillary rise

WT ._ ....... (+) Pore pressure

Figure 2.10. Suction profile (Klausner, 199I).)))

 Effective Stress and Pore Pressurein Saturated Soils 63

Stress (kPa)
-30 -10 10 30 50 70 90 110130150

0.0 m
p = 1.88 Mg/m
3
Silt:
G s = 2.65
w = 10.5%

Dlo= 0\"02 mm
WT
3.8 m \177

5.0 m
\177
6
G s= 2.7, w = 35%
Clay: Clay
7
8.0 m 8
Pore pressure Effective
(a)
(b)

Figure 2. I 1. Example 2.5.

Example 2.5

For the soil profile shown in Figure 2.11(a)

plot the distribution of total and effective ver-
tical stressesand pore pressure taking account of the capillary rise in the silt layer.

Solution:

Calculate the dry and saturated densities of silt: 9d = 1.88(1

+ 0.105) = 1.701 Mg/m 3.
3
Assuming V= 1 m :

Vs
= 1.701 / 2.65 = 0.642 m \177,Vv
= 1.0 - 0.642 = 0.358m 3,

Osat
= 1.701 + 0.358x 1.0= 2.059 Mg/m 3.

Calculate equivalent =0.558,

tube diameter: e = 0.358/0,642
D=exDlo
= 0.558x0.02 =0.01116mm, d= l.l16x i0-5 m.
h = -5 = 2.7m.
= 4x0.000074/9,81xl.116\303\22710
c 4T/Ywd
Calculate the saturated density of clay:
s = 0.35\303\227
= wG 2.7 = 0.945,
e(saturated clay)

V
s
= 1/(i + e) = 1/(1+ 0.945)
= 0.514 m \177, V =
v e/(l + e) = 0.945/(1
+ 0.945) = 0.486 m 3,

Psa\177
= 0.514x2.7 + 0,486\303\227i.0 = 1.874
Mg/m 3.
At depth 3.8 - 2.7 = 1.1m:
o v - 1.88\303\2279.81xl .1 =
20.3 kPa, u = -1.0\303\2279.8 ix 2.7 = -26.5 kPa,
crv
= 20.3+ 26.5 = 46.8 kPa.
At depth 3.8 m:
= 74.8kPa, u = 0.0 kPa, o v = 74.8 kPa.
o v = 20.3 + 2.059x9.81\303\2272.7
At depth 5.0 m:
o v = 74.8 + 2.059 x 9.81
\342\200\242
x 1.2= 99.1 kPa, u = 1.0 \303\227 1.2 = 11.8 kPa,
9.81 \303\227
= 87.3 kPa.
v =99.1-11.8
o\"

At depth 8.0 m:
ov = 99.1 + 1.874\303\2279.81\303\2273.0 = 154.2 kPa, u = 1.0x9.81\303\2274.2 = 41.2 kPa,)))

IL
64 Soil Mechanics: Basic Concepts and Engineering Applications

Pressure
measurement

Valve
(a) (b) (c)

Figure 2,12. Conceptual model for consolidation.

= 154.2 - 41.2 = 113.0kPa,

\177v

The results are plotted in Figure 2.1 l(b).

2.3 STATE OF STRESSAT A POINT DUE TO EXTERNAL FORCES

2.3.i Conceptual model for consolidation

External loading increases the total stress at every point in a saturated soil above its initial
value. The magnitude of this increase depends mostly on the location of the point and is

estimated from appropriate stress distribution theories based on the mechanical properties
of the saturated soil (Chapter 5). In clay soils immediately after the application of the load,
the pore pressure increases to almost equal the applied load and the pressurised water
tends to move towards the free drained boundaries. Thus, over time, the excess pore pres-
sure dissipatesand the pore pressure approaches its initial value. The volume of the clay
layer decreases to balance the volume of the transported water and remains saturated.
Thus the applied load is gradually transmitted to the solid particles and the associated vol-
ume change causesthe settlement of the layer. This phenomenon, known as consolidation,

can be explained by the mathematical-physical model shown in Figure 2.12.

Consider a cylinder with a spring supporting a piston and filled with water to the top of
the spring, as shown in Figure 2.12(a). The piston seals the water but has smooth contact
with the wails of the cylinder. Initially a valve, which is the only escape for the water, is

closed and the system is in equilibrium. There is no force in the spring and no pressure in
the water. If a normal force N is applied to the piston the water will resist the entire force
since, because it is incompressible, there will be no deformation in the water and conse-

quently no deformation and therefore no force in the spring. If the valve were to be opened
for a specified period of time, the pressurised water drains through the valve (Figure
2.12(b)). As a resuIt, the spring undergoes a shortening equal to the vertical displacement
of the piston and the load N is resisted by both the water and the spring. If the valve re-
mains open until the water pressure decreasesto zero (Figure 2.12(c)) the spring takes the
entire force applied to the piston. The time over which the force N is transmitted from the
water to the spring depends on the diameter of the valve, the volume of water inside the)))
 Effective Stress and Pore Pressure in Saturated Soils 65

cylinder and the elastic characteristics of the spring. In a real soil the valve represents the
voids between the solid particles and the spring represents the solid particles themselves
and the load N is normally applied through a footing or a similar type of structure. The
pore pressure at every point within the volume increases initially to a maximum value
equal to the increment of stress at that point caused by the applied loading (excess pore
pressure). A gradual process of redistribution of load to the particles starts immediately
after application of the external load (consolidation).The process is associated with vol-
ume change (equal to the drained water) and surface settlement. At the end of this process
the applied load is resisted totally by the soil particles. In unloading the volume change
due to consolidation is not recoverable or the rate of expansionis very small in compari-
son to the rate of compression. Consequently, the conceptual model shown in Figure 2.12
has to modified
be to represent this behaviour.
rate at which the total load is applied also affectsthe
The soil deformation behaviour. If
the external load is applied in small increments over a long period of time and the soil has
free draining boundaries, then there will be no excess pore pressure and the applied incre-
ments of load will by the solid particles. Volume
be resisted change will occur in incre-
ments similar to the loading. In general this type of the loading, whether in the field or in
the laboratory, is termed drained loading. In undrained loading the water in the voids
cannot drain to the flee boundaries and volume remain unchanged. However, in the field,
the lateral boundaries are not as rigid as the walls of the cylinder model, and an element of
soil will undergo vertical and horizontal deformations most probably in the reverse direc-
tions to keep the volume unchanged. The term undrained loading can also be used to de-
scribe the state of stress in the field when the load is applied very quickly and the seepage
of water from the voids to the free boundaries takes place slowly, In laboratory techniques
the slow application of load is assumed to be equivalent to drained loading whilst rapid
load application represents undrained loading. However, the full undrained loading condi-
tion can be created regardless of the rate of load application.
The compression of a clay layer is drawn downwards as a
may occur if the water table
result of pumping. In this case the due to the reduction
effective in the
stresses increase
level of the water table. However, if the pumping is stopped, the water table will recover
to its initial position after time but most of the settlement that occurred in the clay layer
will not be recovered. This forms the basis of a method of soil stabilization to minimize
the volume change due to external loading.

Example 2.6

A layer of clay of 4 m thick is overlain by a sand layer of 5 m, the top of which is the

ground surface. The clay overlays an impermeable stratum. Initially the water table is at
the ground surface but it is lowered 4 metres by pumping. Calculate o\"v at the top and
base of the clay layer before and after pumping.
For sand e = 0.45,Gs = 2.6,Sr (sand, after pumping)
= 50%. For clay e = 1.0,Gs = 2.'7.

Solution:
Calculate the density of sand at Sr = 50% and 100%:
3
For V= 1 m ,
V
s
= 1/(1 +
e) = 1/(1+ 0.45)
= 0.690 m 3,
3
Vu
=
el(1 + e) = 0.45/(1+ 0.45)= 0.31m .
= 0.690 2.6
\303\227 + 0.310x 0.5 \303\227
1.0 = 1.949 Mg/m 3.)))
[350%sat
66 Soil Mechanics: Basic Concepts and Engineering Applications

\177),sat
\177\"
0.690\303\2272.6+0.310\303\2271.0 = 2.104 Mg/m 3.
Using similar calculation for e = i, t
\1773sa (clay)
= 1.850 Mg/m 3.
At the top of the ctay layer before pumping:
= 2.104\303\2279.81\303\2275.0 = 103.2 kPa,
\177v

u = 1.0\303\2279.81\303\2275.0 = 49.0kPa,

= 103.2 - 49.0 = 54.2kPa.

\177

At the base of the clay layer before pumping:

crv
= 103.2 = t75.8 kPa.
+ 1.850\303\2279.81x4.0
tt = 1.0\303\2279.81\303\2279.0=88.3 kPa,
= i75.8 - 88.3 = 87.5kPa.
(y\177

At the top of the clay layer after pumping:

= 1.949\303\2279.81\303\2274.0+ 2.104\303\2279.81\303\2271.0 = 97.1 kPa,
crv

,t = 1.0\303\2279.81\303\2271.0 = 9.8 kPa,

cy = 97.1-9.8 = 87.3 kPa.
v

At the base of the clay layer after pumping:

= 97.1 + 1.850\303\2279.8i\303\2274.0 = 169.7 kPa,
crv
u = 1.0\303\2279.81\303\2275.0 = 49.0 kPa,
\177
= 169.7 -49.0 = 120.7 kPa.

Ao-'v (at the top) = 87.3 - 54.2= 33.1

kPa.

Ao\"v (at the base) = 120.7- 87.5= 33.2kPa.

The increase in the effective vertical stress throughout the clay layer is uniform.

2.3.2 The case of partiatly saturated soil

Bishop (1959) extended Equation 2.1 to represent the state of stressin an unsaturated soil:

cr = or' + ua - )t(u
-
(2.8)
a u\177,)

where Ua is the pore air pressure, Uw is the pore water pressure and Z is a constitutive pa-
-
rameter depending on the degree of saturation. Note that (Ua u\177,,) represents the soit suc-
tion.

To explain the physical meaning of Z we rearrange the terms in Equation 2.8:

cr = \177r\"
+Zu w +(1-Z)Ua --\177u =
Zuw + (1-Z)u a (2.9)
It can be seen that Z is the fraction of the unit area occupied by water, and i - Z represents
the area of the air. The magnitude of Z can be determined experimentally, however a lin-
ear relationship between degreeof saturation and Z may be adopted betweenthe following
limits: Sr = 0, Z = 0, Sr = l, Z = 1. As the pore water and pore air are assumed to have an
interface due to surface tension, then the pore water pressure must be always less than the
pore air pressure. Therefore the reliability of Equations 2.9 reduces with a decreasing de-
gree of saturation; however, for higher degrees of saturation which are on the wet side of
the optimum moisture content, Equations 2.9 work well.
Fredlund (1973, 1979)and FredIund & Morgenstern (1977) investigated the stress state
in an unsaturated soil assuming four independent phases viz., solids, pore air, pore water)))
 Effective Stress and Pore Pressure in Saturated Soils 67

and the air-water interface. Three stress variables of \177- Ua,

-
\177Y u,,-, and ua
- Uw, were

introduced to express the state of stress in a three dimensional system of which only two
are independent, as adding the first to the third variable
variable will yield the second
vari-
able. The strength parameters of the unsaturated soil were then defined differently to those
of saturated soils in classical soil mechanics. The conditions of transition to the saturated
case were successfully applied in the proposed models.

2.4 PROBLEMS

2.1 For the soil profile shown in Figure 2.13(a) plot the variation of total vertical stress,
pore pressureand effective vertical stress and indicate their values on the boundaries
of each layer.
Answers: 0% = 31.5 kPa, 58.5 kPa, 92.3 kPa

2.2 For the given soil profile of Figure 2.i3(b) calculatethe effective vertical stress at a
depth of 7.5 m.
Answer: 57.2 kPa
m by a 4 m sand with
2.3 of 4 = 2 =
A clay layer w\177th Psat Mg/m \177soverlain
thtck Psat
3
1.9 Mg/m and Odry = 1.65 Mg/m3, the top of this Iayer being the ground surface.
The water tabte is located 2 m below the ground surface. The clay layer is underlain
by a sand stratum that is in artesian conditions with the water Ievel in a standpipe be-
ing 4 m above the ground surface.
Calculate the effective vertical stresses at the top and the base of the clay layer, If

the dry sand is excavated, in what depth the effective stress at the bottom of the clay
layer will become zero?

Answers: 50.0 kPa, 30.4 kPa, 1.9 m

3
2.4 A clay layer 10 m thick has a density of 1,75 M\177m and is underlain by sand. The
top of the clay is the ground surface. An excavation in the clay layer failed when the

depth of the excavation reached to 6.5 m from the ground surface.

CaIcuIate the depth of water in a standpipe sunk to the sand layer.
Answer: 3.875 m

0.0m \177\177 0.0m \177

= 1.70 3
1.2 WT
Soil 1
Oeq Mg/m
WT w= 14%, Sr = 50%
m \177 1.5 m
0sat= 1 .90 \177-'w\177m3 \177Soill Gs = 2.68
2.5 m
3.0 m
Soil 2: = 2,10 3
psat Mg/m
Soil2: Peat e = 5, Gs = 2.0
5.0 m
= 2.15 3 6.0 m
Soil3: Psat Mg/m
Soil3:e = 1.0, Gs = 2.7
8.0 m (b))))

(a)
Figure 2.13. Problems 2.1 & 2.2.
68 Soil Mechanics: Basic Concepts and Engineering Applications

2.5 REFERENCES

Bell, F.G. 1993.Engineering geology. Oxford: Blackwell Scientific Publications.

Bishop, A.W. 1959. The Principal of effective stress. From a lecture in Oslo, Norway 1955. Re-
printed in Teknisk Ukeblad (39): 859-863.
rd
Fredlund, D.G. 1973. Discussion of the second technical session. Proc. 3 intern, conf on expan-
sive soils 1:71-76.
Fredlund, D.G. 1979. Second Canadian geotechnical colloquium: Appropriate concepts and technol-
ogy for unsaturated Soils, Canadian geotechnicaljournaf (16): 121-139.
Fredlund, D.G. & Morgenstern, N.R. 1977. Stress state variables for unsaturated Soils. Journal
SMFED, ASCE103(GTS): 447-466.
Klausner, Y. 1991. Fundamentals of continuum mechanics of soits. London: Springer-Verlag.
Powrie, W. 1997. Soil mechanics-concepts and applications. London: E & FN Spon.
Terzaghi, K, & Peck, R. B. 1967.Soil mechanics in engineering practice. New York: John Wiley.)))
CHAPTER 3

The Movement of Water through Soil

3.1 INTRODUCTION

This chapter investigates the flow of water through interconnected pores between soil particles
in both one and two dimensions. Often, the flow of ground water through a confined or uncon-
fined porous material can be represented as a uniform flow in a plane. The term seepage flow
is used to describe the movement of water through or under soil structures to wells, drains and
reservoirs. Generally, the Reynolds number defined in fluid mechanics is less than l, and
therefore the flow is laminar and Darcy's law is valid. Any departure from laminar flow is not
serious as long as the Reynolds number is below 10,the upper limit for the validity of Darcy's
law. A simplified theoretical and numerical approach is developed and applied to practical ex-
amples to obtainflow nets that describe the seepage flow. This is followed by descriptions of
test methods (laboratory and in-situ) used to obtain the coef-ficien\302\242 of permeability (or hydrau-
lic conductivity) that controls the velocity and the volume of water flowing per unit of time.

3.2 PRINCIPLESOF FLOW IN POROUS MEDIA

3.2.1 Darcy'slaw
Figure 3.1 shows the flow of water through a tube of soil of length L and cross-sectional area
A. Although water moves through the interconnected voids of the soil, it is convenient to de-
fine the velocity as if the water flows through the whole section. Whilst this imaginary veloc-

ity is many times Iess than the actual velocity, the flow rate calculated using this velocity is
real. The water level at its entry point B into the soil tube is higher than the water level at the
exit C. This shows the dissipation
point of potential energy as the water flows across the
length L. The water level at each point is represented by its distance h from an arbitrary datum
and is called head or total head. According to Bernoulli's definition of head:

It = z +u/y\177 +v 2/2g (3.1)

where z is the head (vertical coordinate of the point),

positional u is the water pressure or pore
pressure, 7w unit weight of water,
is the v is the veIocity and v 2/2g is the velocity head. In
classical soil mechanics, the velocity head term is ignored due to low velocities and therefore:

69)))
70 Soil Mechanics: Basic Concepts and Engineering Applications

Arbitrary datum for h S

Figure 3. I. Idealized diagram for flow of water through a tube of soil.
h = (3.2)
z+u/\177w

The hydraulic gradient within the length L is a dimensionless parameter and is defined as the
rate of change in totaI head (or head loss) Ah over the length L:

i = Ah / L (3.3)
In steady flow conditions the hydraulic gradient along a finite length is assumed constant, oth-
erwise it is a point-related parameter that defines the reduction of total head per unit length of
flow across the specified direction. If the head loss for an increment of/kL is Ah then:

i = Ah / \177 (3.4)

If the total head is expressed as a function of L, then Equation 3.4 is replacedby:

iL = (3.5)
--\177--{-

The negative is in the direction

sign implies that
the velocity of decreasing total head. As long
as the calculations on absolute values and there is no coordinate system involved
are based in
the seepage problem, Equations 3.3 and 3.4 may be used. In two or three-dimensional coordi-
nate systems the hydraulic gradient is defined by Equation 3.5. Darcy, a French hydraulic sci-
th
entist of 19 century, stated that the flow rate through a porous medium is directly propor-
tional to the head loss and inversely proportional to the length of flow path. This may be
written as:

v = ki (3.6)
where k is the coefficient of permeability of the material and is determined experimentally.
The quantity of water that flows in a unit of time orflow rate is then given by:

q = Av = Aki (3.7)

The coefficient of permeability has the dimension of length per unit time and decreases as the
particle size decreases. An estimate of the actual velocity vs, referred to as the seepage veloc-
ity, can be made by considering the following equation:)))
 The Movement of Water through Soil 71

Figure 3.2. The concept of seepage pressure.

q = Av =
Avv s

where Av is the cross-sectional area of the pore voids. Multiplying the both sides of this equa-
tion by length L we have:
ALv =
AvLv s -->
Vxv=VvXV s --> v s =(V /Vv)v=v/n or:

vs
= v/ n = v(l + e)/ e (3.8)
where n is the porosity and e is the void ratio of the soil.

3.2.2 Seepage pressure and critical hydraulic gradient

In an upward one-dimensional flow the effective verticat stress at the base of a column of soil
reducesby an amount equa\177 to 5'w Ah. Figure 3.2 shows an upward flow through a soil column
due to a higher head in the underlying soil. Using Equation 2.1, the effective vertical stress at
the base of the soil column is:

= -
=\177r
\177r\177, v -u (ysa\177L
+ \177whw)
Yw(L + h\177 + Ah)

\177r' = cr - u \177- -
)L
- (3.9)
v v ('tsar \177,\177 \177wAh

If the water level at the column base were equal to the water level at the top of the column,
there would be no upward movement of water and \177v would be equal to the first term of
Equation 3.9. Consideringa soil column of unit area the reduction in o\"v per unit volume or
seepage pressure is:

j = yw\177/L =
iy\177 (3.10)

where i is the hydraulic gradient. The first term of Equation 3.9 represents the submerged
weight of the soil column of length L with unit cross-sectional area. When the seepage pres-)))
72 Soil Mecl\177anics: Basic Concepts and Engineering Applications

sure multiplied by the volume of the soil column becomes equal to the submerged weight of
the soil column, o\"v at the base of the column reduces to zero. The soil particles loose their
contact and the column is subjected to heave and consequently fails. This state is called the

quick condition and defines a critical hydrauIic gradient given by:

=
\302\243r],L(Ysa t -Yw) -Tw zSjt = O, ic = (Ysat -Yw) / Y\177
=
T'/Yw (3.11)

Substituting y' from Equation 1.21 we have:

ic = T\"/'l = - 1)/(1+ e) (3.12)

w \"l w (Gs

It should be noted
that, at zero effective vertical stress, the strength of the soil depends entirely
on its cohesion and therefore the quick conditions for clays may not obey Equation 3.12. In
sands the strength and stability of an element is provided by the forces at the contact points of

the particles. Thus, in the or% = 0 condition, the sand element is no longer stable. The state of
sand in this condition is similar to that of a slurry liquid with a density greater than that of wa-
ter and is sometimes referred to as quicksand. In the case of seepage in the direction of gravity,
it may be shown that the effective vertical stress increases from its static value by an amount
equal to y,\177.Ah. In general the seepage force acts in the direction of flow and may increase or
decrease cr'v depending on the angle between the direction of flow and gravity.

3.3 PERMEABILITY

3.3.1 Coefficient of permeability

The coefficient of k is determined by experimental methods and varies from 1
permeabiIit\177
m/s in gravels to less than 10 m/s in clays (Table 3.1), The magnitude ofk depends on the
size of the particles, their shape and orientations, and the structure of the soil, The viscosity of
the water also has some effect on k as it varies with change in temperature. Whilst many
mathematical expressions have been proposed for the coefficient of permeability in terms of
the physical properties of soils,experimental determination of k is always recommended.

Table 3.1. Range of values for the coefficient of permeability.

k(rn/s)

Very fine sands, organic and [

Clean Clean gravels,clean mixture
gravels sand and inorganic silts, of sand, I Clays
gravel
silt and clay
]

Poody Practically
Well drained soils
drained soils impervious)))
I
 The Movement of Water through Soit 73

WatTupP]Y

Constant lever
of water
water to Jill

-\177&h stand pipe

\177
\177 , T
iCoc\177

\177 \177

Cylinder for
Constant
!l\"\177\"v\302\260Iume \177 \177
level of water
measurement
(a) (b)

Figure 3.3. Laboratory permeability Jests: (a) constant head test, (b) falling head test.

An empirical equation proposed by Hazen a century ago relates the coefficient of permeability
of fine graded granular materials to their particle size distribution characteristics by:
k = CD\177o (3.13)

where k is in m/s and D1o = diameter corresponding to 10% finer in nun and C is a coefficient
that varies from 0.005 for silt and well graded sands to 0.012 for uniform sands.

3.3.2 Laboratory permeability tests

Two tests commonly conducted in the laboratory are the constant head test and the falling
head test.

Constant head test. A diagrammatic view of the test arrangement is shown in Figure 3.3(a)_.
This test \177smost suitable for granular soils (gravels and sands) where/\177 is smaller than 10
m/s. The soil specimen is formed in a cylindrical mould (75 or 100mm diameter) and water is
passed through the sample from a constant head tank. Two outlets, or pressure take-off points,
A and B separated vertically by a distance L (=100 ram) show the water levels at these outlets.
Longer permeability cells with 3 or 6 outlets are also used.Two filters using coarse-grained
soils are placed at the top and the bottom of the cell to ensure uniform flow across the soil
sample. Before any readings are taken, a specified time must elapse to ensure steady flow con-
ditions are achieved, The discharged water Q is collected and measured by one of two meth-
ods, either the volume overa given time t or the time for a nominated volume. The difference
\177 between the water levels at A and B is recorded, and the rate of flow is the ratio of dis-
charged water Q to the elapsed time t. Using Equation 3.7:

q
= --Q = Av = Aki = Ak--,Ah and hence
t L)))
74 Soil Mechanics: Basic Concepts and Enghteerhtg Applications

QL _ qL
k- (3.14)
Ah\303\227At Ah\303\227A

To ensure Iaminar flow conditions a hydraulic gradient of 0.2 to (}.3for loosematerial and 0.3
to 1.0 for dense material is used. The constant head is approximately i20 mm above the top
surface of the sample and is applied through the base of the cell to ensure full saturation.

Falling head test. This method is used for fine-grained soils such as silts and clays with a coef-
ficient of permeabthty between 10 and 10 m/s. A schematic v\177ew of the apparatus \177sshown

in Figure 3.3(b). The permeability cell is a cylinder of approximateIy 100 mm in diameter and
a height of i30 ram. A filter Iayer of coarse-grained materiai is positioned at the top of the
sample to ensure unifo\177 laminar flow. A pefforat\177 disc \177d a porous plate with a wire gauze
rest at the base of the cell and \177e held between the top and bottom plates of the cell by retain-
ing ties. A capilla\177 tube is fitted to the top of the celt and the assembly is placed in a soaking
tank. If the constant water level at the tank is lower \177an the height of the cylinder, precautions
have to be made to ensure full saturation of the soiI specimen in the cell. At the st\177t of the test
and at t = 0 the head is equal to h\177. The valve on the ve\177ical capillaff tube is opened and the
time t for the head to fall to h2 is recorded. The hezd h co\177esponding to time t is reduced by
dh due to an increase in time of dr. Applying D\177cy's law and noting the volume of water dis-

charged from the capilla\177 tube is \177ual to the volume of flow through the sample we have:

q = Av = Aki ..... Ak
h
a
dh
L dt

where a is the internal sectional \177ea of the capiil\177y tube and the negative sign shows that the
increase in time causes a decrease in the head. integrating this continuity condition, we have:

\177 d\177 dh aL h\177

= \177 k = 2.3 log (3.15)
for \177dt Jh,\" (-\177) At h2

Recorded data now are substituted into Equation 3.15 to dete\177ine the coefficient of perme-
ability. To ensure optimum accuracy, various standpipe sizes can be used. Soils of very tow

permeability are seal\177 inside the permeability cell (using a thin layer of wax to the internal
sides of the mould) to prevent seepage and piling of water between the mould and specimen.
Details of both techniques may be found in AS\177 D-2434-68 (2000), BS 1377 and AS

1289.6.7.1 & 2 (1999). Some impogant notes with reg\177d to sample prep\177ation, surch\177ge
and calculations areas follows:
t. Remoulded specimens are compacted with a specified compactive effo\177 and must represent
the field condition. The size of the pe\177eability eel[ depends on \177e maximum p\177icle size of
the soil and is selected according to code r\177uirements.
2. If necessa\177 a confining \177ial load or surcharge is appli\177 to \177e sample to represent the
field stress conditiot\177s.
3. To include the effect of temperature the following equation may be used:
= (3.16))))
kr k0(\17720/\1770)
 The Movement of Water through Soil 75

where 0 is the temperature of the outflow water in degrees Celsius, kT and k0 are the coeffi-
cients of permeability at 20\302\260Cand at O\302\260C,
r12o and rl0 are the dynamic viscosities of water at
and
20\302\260C 0\302\260C
respectively.

-9
Constant head method using aflexible wall permeameter. For soils with k < 10 m/s, a con-
stant head method,
incorporating a flexible wall permeameter, is used (ASTM D-5084 (2000),
BS 1377, AS 1289.6.7.3 (1999)). In this test remoulded or undisturbed specimens are tested
under a specifiedeffective stress to simuIate the field condition. A cylindrical soil specimen,
protected by a thick rubber membrane, is tested in a triaxial cell (Chapter 4). Thecell pressure
is increased gradually causing both the effective stress and the pore pressure to increase within
the sample. A back pore pressureis applied through the top of the specimen and the pore pres-
sure is monitored at the base of the specimen to ensure that the effective stressdoesnot exceed
the required final value. Both the back pore pressureand ceil pressure are increased incremen-
tally until the required cell pressure and effective stress are achieved. At this stage the base of
the specimen is under constant head and the fall in the head at the top of the specimen can be
measured to establish the head loss. The rate of flow is measured over a sufficient period of
time, until the permeability calculated using Equation 3.14 is constant.

Permeability test in Rowe type consolidation cell. A diagrammatic section of a Rowe type
consolidation cell is shown in Figure 3.4 and is the development of an original design by Pro-
fessor P W Rowe at Manchester University UK. Whilst the apparatus is designed for a con-
solidation test, it can to determine the coefficient
be used of permeability for both vertical and
radial flow. The test is carried out under a specified effectivevertical stress but the effective
horizontal stress, which is the reaction of the rigid wall to the applied effectivevertical stress,
is unknown. This, in many ways. simulates the consolidation processunder structures build
upon saturated clay soils. Thecell is commercially available in three diameters of 75 ram, 150
mm and 250 mm and complies with BS 1377. It is manufactured from aluminium alloy and

consists of a cylinder, and top and base plates, The soil specimen within the ceil is hydrauli-
cally Ioaded through a diaphragm that ensures a uniform distribution of total vertical stress at
the top surface of the sample. Under a specified effectivevertical stress different drainage
conditions vertically upwards or downwards or radially outwards or inwards can be applied.

Diaphragm D []
Back
pressure pressure
system system

--
Water

Sample
\177
..............
\177 P Por\177 pressure
\177 measuring system
\177x.\\\177-,.\\\\,..\177

Figure 3.4. Rowe type hydraulic consolidation cell.)))

76 Soil Mechanics: Basic Concepts and Engineering Applications

Figure 3.5. Example 3. I.

Permeability are carried out in the 250 mm diameter cells. Both

tests methods of constant head
and falling can be performed depending on the type of soil. In the vertical permeability
head
test two porous drainage discs are placedat the top and base of the sample. For radial flow po-
rous plastic liners are placed around the sample with a central porous hole filled with sand.
The diameter of the hole is approximately 5% of the sample diameter. Categories of the sam-
ple include undisturbed, remoulded, compacted soil and undisturbed prepared from remoulded
specimens. Reference may be made to Head (1986) for a detailed test procedure.

Example 3.1

An impermeable dam is constructed on 0.5 m thick lay, er of silty sand as shown in Figure 3.5.
3 -4
Estimate the flow rate per metre width of the dam in m/day, k = 2 \303\227
10 m/s.

Solution:
Assume a constant gradient along the silty-sand
hydraulic layer: i = Ah I

Assume the toe dam as the datum for total heads:

of the
= 8.0 + 2.0 = 10.0m.
h 1 = total head at the
upstream side = 8.0+ 40.0 \303\227 (1/20)

h2 = total head at the downstream side = 2.0 m.

Ah = h - h = 10.0- 2,0= 8.0,L --- \177/40.02 + 2.02 = 40.05, i = bJ\177/L = 8.0/40.05 = 0.1997.
1 2
q = Aki
-4 - 1.73 m3/day.
= (0.5\303\2271.0)\303\2272.0\303\227i0 24
\303\2270.i997\303\2273600\303\227

Example 3.2
A head permeability
constant cell has an internal diameter of 75 trim and three tapped outlets
A, B,C at a 100 mm vertical
and pitch. During three sets of tests on a specimen of sand the
following data were recorded. The flow was upward and each test was conducted with a dif-

ferent constant head. Determinek in mJs (average of six values).

Test Q (mill0 minu[es) Head (ram) abovedatum

no. A B C

1 281.5 139.0 100.0 60.0

2 422.5 208.5 150.5 91.5
3 591.0 292.0 210.0 127.0

Solution:
A= area of cell= 75.02
x\177/4xl0
-6 = 4.4179x 10 -3 m 2. Using Equation 3.14 for each test:)))
 The Movement of Water through Soil 77

Testl: q = 281.5/10.0= 28.15ml/min =28.15\303\22710 -6/60.0 = 0.469\303\22710 -6 m3/s,

(a) k = 0.469\303\22710
-6
\303\2270,1/[(0.139- 0.10\1773)\303\2274.4179\303\227i0
-3
]
= 2.72\303\22710 -4 m/s,
-6 -3 =4
(b) k =0.469\303\22710 \303\2270.1/[(0.100-0.060)\303\2274.4179\303\22710 ] -- 2.65\303\22710 reds.
Test 2: q = 422.5/ 10.0= 42.25ml/min = 42.25 \303\22710 -6/60.0 = 0.704\303\227 10 -6 m3/s,
-6 -3 =
(a) k = 0.704\303\22710 \303\2270,1/[(0.2085- 0.1505) \303\2274.4179 \303\227 l0 ] 2.75 \303\22710-4m/s,
- -3 =
(b) k = 0.704\303\227 10-6 \303\2270.1/[(0.1505 0.0915) \303\2274.4179\303\227 10 ] 2.70\303\22710-4 m/s.
Test 3: q = 591.0/10.0= 59.10
ml/min = 59.i0\303\22710 -6/60.0 = 0.985\303\22710
-6
m3/s,
-6 -3 -4
(a) k = 0.985\303\22710 \303\2270.1/[(0.292-0.210) \303\2274.4179\303\22710 ] = 2.72\303\22710 m/s,
= -6 -3
(b) k 0.985\303\22710 \303\2270.1/[(0.210 -0.127)\303\2274.4179\303\22710 ] = 2.68\303\22710-4 rrds.
-'\177
Average k = 2.70 10 m/s.
\303\227

Example 3.3

A falling head was conducted in a cell of 100 mm diameter and 127 mm high. The re-
test
corded data were: hi =560 ram, h2=465 ram, t = 1081s, and the diameter of the vertical
standpipe was 7 ram. Determine (a) k in m/s, (b) the time required for a similar head and drop
if a standpipe 4 mm in diameter was used. Include the effect of capillary rise.

Solution:
= 0.12 = 7.854\303\22710 -3 = 0.0072 \303\227\177z/4
= 3.848\303\22710 _5
(a) A \303\2277\177/4 m 2, a m 2,
From Equation 3.15:
k = 2.3 \303\2273.848\303\227105 \303\2270.127/(7.854\303\227 10 -3 \303\227
t08 t.0)
-7
= 1.069\303\22710 m/s.
1og(560.0/465.0)
Correction for capillary rise: he = 4T/Ywd = 4\303\2270.000074/9.81\303\2270.007\303\2271000 = 4.3 rnm.
hi = 560.0 -4.3 = 555.7ram, h2 = 465.0 -4.3 = 460.7ram.
-7
= 1.078\303\227
k = 2.3 \303\227
3.848 10-5 \303\227
\303\227 0.127/(7.854 10-3 \303\227
\303\227 1081.0) 1o g(555.7 /460.7) 10 m/s.
=
(b) c 4T/\302\245wd
h = 4\303\2270.000074/9.81\303\2270.004\303\2271000 = 7.5 ram,

-
hi = 560.0 7.5= 552.5 trim, h2 = 465.0 - 7.5 ram. =457.5
-5
Ford =4 mm, a = 1.257\303\227
10 m 2. Using 3.15:
Equation
-5 -3 -7
t = 2.3\303\2271.257\303\22710 \303\2271.069\303\22710 ) =
\303\2270.127\303\2271og(552.5/457.5)/(7.854\303\22710 358 s.
From Equation 3.15 it can be shown that in the absence of capillary rise:

t I ] t2 = = 7.02 / 4.02 = 353

D? / D\177, 1081,0/t 2 \177x
t2 s.

Example 3.4
A permeability test is arranged according to Figure 3.6. The cell has an inner diameter of 100
mm and length of 300 rnm. Two soil specimens of equal length
a total are fitted into the cell as

shown. Sotl 1 has a coeff\177clent of permeablhty of 6,5 x 10 rrds. The total heads at points A
and B (hA and hB) have been recorded650 mm and 320 mm respectively. The amount of

discharged water is 210 ml in 5 minutes.Calculate the coefficient of permeability of soil 2.

Solution:
-3
A = 7.854 10
\303\227 m 2, L= 0.3/ 2 = 0.15m, using Equation 3,14 for soil 1:
= -6 -5 -3 =
Ah = QL/kAt 210.0\303\22710 \303\2270.15/(6.5 \303\227t0 \303\2277.854\303\22710 \303\2275.0\303\227
60) 0.2057 m.
For soil 2: Ah =0.65 - 0.2057 - 0.32= 0.1243m,
= -3 -4)))
k = QL/AhAt 210.0\303\22710-6 \303\2270.15/(0.1243\303\2277.854\303\22710
=
\303\2275.0\303\22760) 1.075\303\22710
78 Soil Mechanics: Basic Concepts and Engineering Applications

Soil 1
B
.... q
\177\"

L \177 L

Figure 3.6. Example 3.4.

3.3.3 In-situ permeability tests

The main disadvantage of a laboratory is the small size of the
permeability test sample, It is

inappropriate to use
only laboratory results to study
conditions
flow in the field. These results
must be supported by in-situ tests before selectingan average k for the conditions involved in
the project. In-situ permeability tests require the construction of test and observation wells
both of which are expensive. Moreover, the complete hydraulic behaviour of any aquifer may
need to be known. A comprehensive pumping test provides the necessary information about
permeability characteristics and is a reliable basis for the prediction of ground water behav-
iour. In-situ tests are nortnally based either on the constant or variable head methods. The flow
rate and the position of the piezometric level in the aquifer must also be determined which
could be unconfined, confined or a combination of both. Care has to be taken to avoid distur-
bance, which may happen if aquifers are connected through the test wells and boreholes. A
typical section of a modern test well is shown in Figure 3.7 (AS 2368). The well is sometimes
protectedagainst instability of its sides by casing but in dense and consolidated soils there
may not be a need for casing. Water is extracted from the screened interval(s) that may also be
perforated or unlined.

Pumping test in an unconfined aquifer. An idealized section of an unconfined aquifer with a

pumping well and two observation wells is shown in Figure 3.8. A pump is selected in accor-
dance with the relevant to the maximum discharge and is mounted
code with
regard at the top
of the pumping well to discharge a specifiedflow rate of q. As a result of pumping, the water
level in the well is lowered (drawdown) creating a localized hydraulic gradient as shown in

Figure 3.8. This causes the water to flow into the well in an axisymmetric condition making a
cone type potentiometric surface around the well. The shape of the inverted cone (cone of de-
pression)dependson the quantity of the discharge and hydraulic properties of the aquifer. Af-
ter equilibrium is reached, the water levels in the two observation wells are measured using
manual, automatic, electro,lic recorders or pressure gaugesand the total heads of Zl and z2 are
established. The flow passing through any cylindrical soil volume of radius r and height of z is

equal to the discharged flow:)))

 The Movement of Water through Soil 79

Pressure gauge, air valve Piezomelric tube

and sampte tap
Orifice plate

\177Weir
Electric motor drive
tank

Datum

Surface casing \177Cement

grout

\177ump column

Well casing drive shaft

_ _W_a_te E I_e.vel
Non-retum
Drawdown Pump bowls
measurement tube intake
Well screen

Aquifer Aquifer

Sump

Figure 3.7. Typica[ test well in an unconfined aquifer.

q = Av = Aki
=(27trz)k(\177-\177--), qdr= 2\177zdz
dr r
2 dr = 2gk --> k - q r2
qfr fz\177.
zdz In (3,17)

From Equation 3.17 it can be seen that for multiple observation wells with a constant flow rate

\177 r I
Ground surface

\177 wel
WT r2

well \177bse\177atio\177
Pumping
[\177 ,d aquifer
Imp \177rmeable z\177nConfin Zw

\177

Figure 3.8. Pumping well in an unconfined aquifer.)))

80 Soil Mechanics: Basic Concepts and Engineering Applications

9
the ratio of a = ln(r 2 /r l) to b = (z\177 - z t2 ) remains co\177stant for each pair of observation wells.
This implies that, in a coordinate system representing a and b, the test data must be averaged

by a line of best fit. The slope of this line with b axis gives the best average value for a/b to be
used in Equation 3.17. Rearranging this equation, the piezometric level z, in terms of r and one
of observation well data, may be expressed in the following form:

z = \177z\177+(q/\177zk)ln(r/r\177) (3.18)

The radius of influence ro defines the point(s) where the piezometric level fuIly recovers to its

original value of Zw. Substituting zw in Equation 3.i8, ro can be expressedas:

ro =
q exp[\177(z\177, -z\177)/ql (3.19)

Accordingly, the theoretical drawdown in the well is obtained by replacing rw for r in Equa-
tion 3.18 and subtracting the result from the original piezometric level Zw.

D w = Zw -
\177[Z?
+ (q/Tr.k)In(r
w/ rt) (3.20)
Note that in the application of the Darcy's law we assumed that the water moves horizontally
towards the centreline of the well and perpendicular to the soil cylinder the two ob-
between
servation wells. Thus vertical components of the velocity, which may be of significance close
to the well, are ignored. Furthermore, due to the head loss caused by screens and gravel packs
in the bottom of the well, the observed water level in the well may be lower than the water
level immediately behind the well easing. Correction factors for the drawdown in the test well
are available from standard codes (e.g.BS 1377). These reduce the actual drawdown to a more
realisticvalue which is used to assess the overall behaviour of the aquifer.

Pumping test in a confined aquifer. SimilarIy, for a confined aquifer of thickness D (Figure
3.9) subjected to a discharge flow of q, Darcy's law can be applied to obtain k:

dr r2 dr
Av= Aki=(27trD)k( )--) = 2\177&Ddz --> q[ = dz
q= 27\177kD[r.z
\177rr q---\177- \177r\177r \177z\177

\177
rl
q
\177-I
\"\177--\177 Ground surface

Piezometric level

=ira,ore= I * aquifer
t

Figure 3.9. Pumping well in a confined aquifer.)))

 The Movement of Water through Sol1 81

k = q In(r 2 / q )/[2\177D(z - z )] (3.21)

2 t

By rearranging Equation 3.21, the piezometric level, radius of influence and drawdown in the
well are expressed by Equations 3.22 to 3.24 respectively.
z = zt + (q/2gDk) In(r/r 1) (3.22)

ro = q exp[(2\177tDk/q)(zw - z
I )] (3:23)

D\177
=
zw -[q + (q/2\177zDk)ln(r\177. / q )] (3.24)
\177
tn both unconfined and confined aquifers to estimate the radius
il is useful of influence ro
where the localized water level recovers original level
to (z =
its z\177). The drawdown in the
well may be approximated from Equations 3.20 and 3.24 without the need for an observation
well if an arbitrary value of ro = 300 m to 350 m is substituted in these equations (Roberson et
al., 1997). However, the result must be interpreted with care and considered only as a prelimi-
nary step towards.the hydraulic study of the aquifer, Another approximation reported by
Linsley et al. (1992)is to assume that the radius of influence is proportional to the flow rate. In

this case, by substituting two sets of data comprising two flow rates and two drawdowns into

Equations 3,20 or 3.24, estimate

an of the coefficient of permeability
and drawdowns for spe-
cific flow rates can be obtained. In confined aquifer if, as a result of pumping, the water level
is lowered to the aquifer zone, then a modification in the flow rate equation is necessary as the
flow in zone adjacent to the well transfers to the unconfined state,
Equation 3.21 may also be applied to the Rowe type cell for the determination of the radial
coefficient of permeability. In this case D is the height of the sample and r2 and rt represent
the radii of the sample and the central hole respectively. The term z2 - z\177is replaced by \1773p/
13,\177gwhere Ap is the recorded pressuredifference between the porous boundary and the central
hole, 0,\177 is the density of water and g = 9.81 m2/s.

Conventional slug test, This in-situ test is described in ASTM I3-4044 (1996) and AS 2368
(1990) and is categorized as a rapid displacement method. The test has achieved widespread

Figure 3AO, Conventional slug test.)))

82 Soil Mechanics: Basic Concepts and Engineering Applications

acceptance due to its simplicity, as there is neither a need for a pump installation or for obser-
vation wells. The test studies the rapid displacement of the water level in the well by instanta-
neous removal or injection of water. In a conventional slug test the water within the well is

rapidly removed by bailing and is then allowed to recover. Records of the response of the well
are obtained by measuring the recovered water level at different intervals of time. Figure 3.i0
shows two positions of the recovered water level in a welt over a time interval of t.

Auger hole test. This test is similar to the slug test but the removal of water is carried out in an
unlined cylindrical hole drilled with an auger. The results may be affectedby the instability of
the hole's internal walls. However, problems due to well screen and gravel pack that influence
the slug test results are non-existent.

Piezometer test.test is similar to auger

This hole test with the difference that a piezometer (a
rigid pipe [ i0 mm in diameter)
approximately is inserted into the aquifer. Advantages of this
method over the auger hole test include the possibility of testing a particular zone of the aqui-
fer and improved stability of the hole.

Pneumatic displacement test. In this test drawdown in the water Ievel ira the well is achieved
by increasing the air pressure above the water. After sealing the top of the we\177l, the air pres-
sure is increased by means of an air compressor. The volume of water associated with the

downward displacement of the water level is absorbed by the aquifer. Constant air pressure is
maintained until equilibrium is reached. The pressure is then suddenly released and the dis-
placement of the water level is measured at various time intervals as ira the slug test.

Tracer test. In this method the time interval for a trace (a dye or a salt) to travel between two
observation wells is measured. The average velocity is calculated by dividing the distance be-
tween the two observation wells by the measured time. By measuring the water levels in the
test wells, the average hydraulic gradient can then be calculated. The coefficient of permeabil-
ity is the ratio of the velocity to the hydraulic gradient. This result has to be considered as an

approximation due to limitatious in the field as well in the theory. The direction of the flow
has to be established by constructing multiple observation wells, which in turn increases the
cost of the operation. Moreover, the existence of low or high permeability layers or of lenses
may affect the resutts. For more information about the pumping tests and investigation of the
aquifer response, reference may be made to BS 5930 (198i) and AS 2368 (i990).

Example 3.5
A
pump test was carried unconfined layer of sand 15 m thick and overlying an im-
out in an

permeable stratum. table was 2,m below the ground

The water surface. The discharge after
achieving equilibrium was measured 1.1 mQmin whilst the drawdown in the two observation
wells at 20 m and 40 m from the pumping well were measured at 0,80 m and 0.55 m respec-

tively. Estimate the average k and ro.

Solution:)))
 The Movement of Water through Soil 83

rl =20,0 \177, zl = 15.0 - 2.0 - 0.80= 12.2 m, r2 = 40.0 m, z2 = 15.0- 2.0- 0.55 = 12.45 m,
q = 1.1m\177/min = 0.018333 m3/s. From Equation 3,17:

- 12.202)] -a
= 6.56\303\227
k = 0,018333\303\227 In(40.0/20.0)/[r\177(12.452 10 n\177s. Using Equation 3.19:
ro = 20,0exp[n\303\2276.56\303\22710 4(13.02 -12.22)/0.018333] = 193 m.
Example 3.6
\342\200\242 6 .
A\177ump
test was c\177ried unconfined aqmfer of k = 3 x t0 \177s w\177th a flow rate of-09
out in an
m\177lhour. The radius of the well is 0,4 m and the aquifer has a depth of 80 m above an
\177mper-
meable stratum. The drawdown in an observation well at a distance of 150m from the well is
2,5 m. Calculate the radius of influence and the depth of water in the well\342\200\242

Solution:
-3
= 77.5 m, q = 20 m3/hour
z\177
= 80,0 - 2.5 r\177
= 150 m, t\177.= 0,4 m, = 5.555 x 10 m3]S.
From Equations 3.19 and 3.20:

tb
=
150.0xexp[\177x 77.52)I5.555x 10-3] = 292.5 m.
3x 10-6 (80.02 -
= -3
Du. 80.0-ff77.52 + 5.555x10 ln(0.4/150.0)/(=x3xt06) = 29.9 m.
Example 3.7
A well is construct\177 to fully penetrate a confin\177 aquifer of thickness of 25 m. The water
level at two observation wells 40 m and 150 m from the well are 1.1 m and 0.4 m below the
original piezometric level respectively.Determine the value of k. q =2.4 m\177/min.
Solution:

q
= 2.4 / 60 = 0.04 m 3 Is, zt = z\177.- 1.1, z2 = Zw- 0.4 \177 z2- = 0.7
z\177 m. From Equation 3,21:

k = 0.041n(150.Oi40.0)l(2nx25.0xO.7) = 4.8x10 4 \177s.

Example 3.8

An unconfined aquifer has a thickness of 40 m. A well of diameter 0.3 m is construct\177 for a

pump test. For flow rates of 43 m3/hour and 135 m3/hour the actual drawdowns in the welt ae

measured 3.7 m and 16.1 m respectively. Estimate the drawdown in the well for a flow rate of
3 .
60 m /hour assuming the radius of influence is propo\177ional to the flow rate.

Solution:

Using Equation 3.20 for tests t and 2:

(a) 3.7 = 40.0- + (43.0/3600.0) In(0.15/t;\177)/(=xk)

\17740.02

16,1 = 40.0- \17740.02

+ (135.0/3600.0) In(0.15 / to2 )/(= x k)
But to2 = (135.0 / 43.0)rol = 3.139rol,thus:
(b) 16.1 = 40.0 - + (t35.0/3600.0) In(0.15/3.139() I )l(\177x k)
\17740.02

Solve two equation, s of (a) and (b) for ro\177 and k:)))
84 Soil Mechanics: Basic Concepts and Engineering Applications

-5
rol = t84.5 m, to2 = 184.5\303\227
3.139 m, k = 9.58 \303\227
= 579.2 10 m/s. For q = 60.0m3/hour:
ro3 = (60.0 / 135.0) \303\227
579.2 or (60.01 43.0) \303\227
184.5 = 257.4 m.

Dw - -5
= 40.0
a]40.02 + (60.0/3600.0)In(0.15/257.4)/(rt\303\227 9.58 \303\227
10 ) = 5,54 m.

3.3.4 Permeability of stratified soils

Consider stratified
the soil section shown in Figure 3. i 1. The averagevalues of k in the x and z

directions are estimated assuming that the water moves either parallel or perpendicular to the
layers. When the seepage is parallel to z:

q = Av = Aki =
Ak:l 1 / z 1)
(z\177J\177
=
Akz2 (Ah 2 / z 2) ..... Akzn (Ah n / Zn )

where Aht, &h2, and \177n are the losses in the total head as the water passes through the layers
of thickness of zb z2 ..... and zn. The total loss is equal to the sum of head losses, hence:
\177. = Zkh + Zkh +-.- + n , or:
Zkh
I 2

\177j\177=qz_\177_\177
qZn q ZI . z2 - qz
+ qz_\177_? +...+\177=___,__\177_...+ ( __zn
)=
Akzl Akz2 Akzn A kzl kz2 kzn
Ak z

where z = zI + z2 + --. + z n and k: is the average for the stratified soil in z direction, thus:

Z
k =
z (3.25)
+
z\177/kzI z2/kz2 +'\"+zn/kzn
Note that the xz coordinate is selected to match the geometry of the layered structure and the z-
axis may not be in the direction of gravity. For the seepage in the x direction the total flow rate
is the sum of flow rates passing through each layer:
q=ql +q2 +'\"+qn, \302\260r:
= A + A2kx2i +.-. + Ankxni
Akxi lkxli
where A = zxl = (z
I
+ z 2 +...+ zn)xl, AI = Zl x 1 ..... and An =zn x 1, Substituting A values
into the equation above, we obtain:
=
(Zlkxl + Z2kx2 +
\"\"
k
x
+
znkxn)] z (3.26)

\177 X

kxl, kzl, z1 1

kx2., k-e2, Z 2 2

kxn ' kzn , zn tl

Figure 3. I I. Analysis of average horizontal and vertical coefficient of permeability in stratified soil.)))
 The Movement of Water through Soil 85

Example 3.9
The profile of a stratified soil contains three horizontal layers each having a thickness of 0.3
-2
m. The top of the first layer is the ground surface. The k values are 10 m/s for the top, 2.5 \303\227
4 -5
10 m]s for the middle and 4.0 \303\227
10 ;rds for the bottom layer. There is 0.2 m of water above
2
the surface of the soil. The sectionis drained at the base level. Calculate the flow rate per m
of the stratified soil in litre/hour. If the overflow mechanism of the constant head is deleted,
what will be the time required for the water level to drop to the ground surface.

Solution:

Using Equation 3.26: = 0.9/(0.3/10 -2 + 0.3/2.5\303\22710 -4 +0.3/4.0\303\22710-5) = 1.03\303\227I0

-4
nfs
kz
-4
q = Aki = 1.0\303\2271.03\303\227i0 >\177(0.2 + 0.9)/0.9\303\2271000\303\2273600=
453.2 labour.
2
Using Equation 3.15 and assuming that, h\177= 1.1 m, h2 = 0.9 rn, L=0.9 m and A =a = 1 rn :
-4 = = 2.3(0.9/t)log(1.1/0.9) --.at
k = 10
1.03\303\227 2.3(aL/At)Iog(h i
/ h 2) = 1751 s = 29 min., 1is.

3.4 FLOW NETS

3.4.1 Flow nets in an isotropic material

A flow net is a graphical representation of and consists of two

two-dimensional
seepage
groups of curves of flow lines lines (Figure 3.12). A flow line in a steady
and equipotential
flow condition represents a curve on which a particle of water cormnences its movement from
a known head (point S, Figure 3.12) and terminates at a specified lower head (point F, Figure
3.12). The tangent at a point on a flow line represents the direction of the velocity defined by
Darcy's law. Thus the velocity at each point is proportional to the hydraulic gradient which in
turn depends on the coordinate of the point. As a result the average velocity on a flow line be-
comes dependent on the length of the flow line. A shorter flow line means a higher hydraulic
gradient for a specified loss of head, resulting in a higher velocity. Referring to Figure 3.12 it
can be seen that the velocity in the vicinity of the base of the dam is much higher than the ve-

locity immediately above the impermeable stratum. An equipotentiat line is a curve on which

Figure 3.12. Flow and equipotential lines under an impermeable dam.)))

86 Soil Mechanics: Basic Concepts and Engineering Applications

the total head is constant (water rises to a constant level) magnitude ranges between
and whose
the highest and lowest total heads relatedto seepage problem. The flow
the lines are every-
where normal to the equipotential lines, and thus the velocity at every point is perpendicular to
the equipotential line passing through that point. From the definition of flow and equipotential
lines we can see that their numbers are infinite, however a finite number of flow and equipo-
tentiat lines (flow net) can give valuable information on seepage characteristics such as veloc-
ity and flow rate. In steady flow there is no flow occurring across the flow lines and the rate of
flow between two flow I[nes is constant. Figure 3.13 is an example of a soil element bounded
with two flow lines and two equipotential lines. The flow net is constructed in a way that the
average distance between two flow tines within an element is equal to the average distance be-
tween two equipotential lines (li in Figure 3.13). Furthermore, the head toss between succes-
sive equipotential lines is equal. The flow lines selected to represent the flow net are those in
which the amount of flow between two pair of flow lines is equal. Two flow lines represent a
lane (Wu, 1966) with a variable section and the total flow rate is the flow rate corresponding
to one lane multiplied by the number of lanes. The flow rate passing through an element abcd
(Figure 3.12) is the constant flow between two flow lines ofqi and qi + 1 and is given by:

zkh -
h2 = \177
Aq = Av =k
= (l = kAh hi
= Aki
i
\303\227 k \303\227
1)\303\227 k
ti N d Nd

where Na is the number of equat drops in totat head and h is the total loss due to the seepage.
If the totat number of flow lanes is Nf
the total flow rate is:

q = NfAq = kh
AIf
(3.27)
Nd
Flow nets are constructed either analytically or by numerical methods. Both techniques are
based on the solution of the continuity equation (Section 3.5). In seepage problems with sim-

ple boundary conditions the flow net is constructed by sketching, as demonstrated in Figure

3.14. This may be done by hand or by computer graphics in accordance with the following:
1. The boundary between soil and water is an equipotential line.

.1I

t \177q

\177\"\"'---t\177
hi t
\177'
hi+ 1

Figure 3.13. Flow and equipotentia] lines.)))

 The Movement of Water through Soil 87

(b)

(d)

Figure 3.14. S\177ep by step construction of a flow net.

2. Any boundary between the soil and an impermeable material is a flow line, The imperme-
able material is either an impermeable stratum(s) of the aquifer or the base of the foundations
and sheet piles that touch the aquifer.
3. Sketching may start by constructing the first flow Iane as shown in Figure 3.14(a). The flow
lane is divided into a number of squares and equipotential lines are projected outwards into the
second flow lane.
4. Thesecondflow line is drawn to form the squares of the second flow lane.
5. The procedure is co,tinued until the last flow lane is formed. This lane must consist of
squares and satisfy the boundary conditions; otherwise the first flow line has to be reposi-
tioned and the whole procedure repeated.

Example 3,10
A sheet piling system with its corresponding flow net is shown in Figure 3.15. (a) Estimate the
flow rate m m /day per 1 m run of p\177hng, (b) for the element A with t = 1.5 m calculate the av-
erage velocity and the effective vertical stress, (c) determine the magnitude of the effective
vertical stress at the base and at the right-hand side of the sheet pile, (d) calculate the factor of
safety against the quick condition, defitmd as the ratio of the existing hydraulic gradient along
the downstream face of sheet \177ile
to the critical hydraulic gradient defined by Equation 3.11.
k = 0.02 ram/s, Ysat
= 20 kN/m-.
Solution:
(a) Nf= 5 and Nd = 11, h = 3.0 m,
q= khNf
/ N a = (0.02 / 1000)\303\227
24 \303\227
3600 5 /
3.0 \303\227
\303\227 11 = 2.356 m3/d ay.

(b) v = ki = 0.02 \303\227

[(3.0 / 11) / 1.5] 3600
\303\227 = 13.1 ram/hour.

L)))
88 Soil Mechanics.\" Basic Concepts and Engineering Applications

stratum

Figure 3.15. Example 3.10.

Head loss at the centre of the element = 2.5

drops x 3.0/ 11= 0.682m,
= of A from water surface - loss of = (4.5 + 1.5) -
head 0.682= 5.318m,
hw depth
u =5\177318\303\2279.8I = 52.2kPa, =4.5\303\22720.0+1.5x9.81 = 104.7 kPa,
\177v
= 104.7 -52.2 = 52.5 kPa.
\177'v

(c) Head loss at the pile

= 6 (number of equal drops)x (3.0/
base of the sheet 1 \177)= 1.636 m,
Height of water at from the water surface - lossof head
this point = depth
= (4.5 + 1.5) - 1.636= 4.364m, u = 4.364x 9.81 = 42.8kPa.
= 3x 20.0 = 60.0 kPa, = 60.0- 42.8 = 17.2 kPa.
\177v \177'v
=
i (4.364-3.0)/3.00.455, i c )\"/)'\177v = (20.0-9.981)/9.81
= = = 1.039.
F = 1.039/0.455-=
2.28.
The factor of safety against the quick condition is also estimated by evaIuation of the stabiIity
of a prism of soil D (embedment depth) by D / 2 (width) by unit thickness attached to the back

of the sheet piIe.The factor of safety is defined as the effective weight of the prism divided by
the upward force resulting from the seepage pressure at the base of the prism.

3.4.2 Flow nets in an anisotropic material

Consideran anisotropic soil with horizontal and vertical coefficients of pcrmeability of kx and

kz respectively. The material can be treated as an isotropic soil by assuming an isotropic

coefficient of permeabilty of:
k =
i \177 (3,28)
and using a transformed scale of:

x =x /k x (3.29)

Considering an element of soil in x direction, having the length of I and cross-sectional area of
the flow rate in the x direction in the untransformed state is: -= = \342\200\242
A, qx Akxi AkxAh/l
In the transformed state:
Ah
= = = A _
qx\" Akii gki k\177-\177xkz _\177--))) x
f l Akx_\177=q
Ah/kx
.cz--\177[l\177z \177[kz\177h/kx
 TheMovement of Water through Soil 89

8.Ore

kx=2.56 X 10-5 m/s

8.0m
kz=1.00 X 10 \1775m/s

impermeable stratum impermeable stratum

(b)

Figure 3.16. Example 3.11.

Examp]e 3.11
A concrete dam has a base length
of 8 m and retains 5 m of water as shown in Figure 3.16(a).
The water downstream side is at the ground surface. Under the
level on the dam there exists an

8 m thick layer of anisotropic permeable soil with the k values shown in the Figure 3.16(a).
Calculate the flow rate under the concrete dam in I/day.
Solution:
-s = 0.625.
The scale factor in the x direction is: \177z/kx
= ffl.0\303\22710 -5/2.56\303\227 10

The length of the transformed base is: 8.0 \303\227

0.625 m. = 5.0
The equivalent isotropic coefficient of permeability is calculated using Equation 3.28:
-5 -5. -5
ki = 10
\177]1.0\303\227 \303\2272.56\303\22710 = 1.6\303\22710 rrds.

Figure 3.16(b) shows the transformed section of the soil-dam, with the flow net drawn by

sketching in which: No, = 8, Nf= 5,

-5
Using Equation 3,27: q = / Nd = 10
1.6\303\227 1000\303\2273600\303\2272418
5.0\303\227
\303\227 = 4320
1/day.
khNf

3.4.3 Control of the flow rate under impermeable dams

The flow rate in the soil under an impermeable dam (Figure 3.17(a)) and the factor of safety
against the quick condition may be controlled by introducing horizontal or vertical imperme-
able blankets and a toe filter. A horizontal impermeable blanket may contain a thin layer of
compacted clay constructed on the upstream side (Figure 3.17(b)). This increases the length of
the flow line inm\177ediately under the dam, projecting all the flow lines to the ]eft of the blanket.
As a result, the number of equipotential lines thereby reducing the flow rate (Equa-
increases,
tion 3.27) to a required value. It may be that
shown if the number of equipotential lines is kept
the same as in the unmodified section, then the number of flow lines and the flow rate will

both decrease. In addition, the pore pressure at the toe decreases,improving the overall stabil-
ity of the dam and tl\177e factor of safety against the quick condition. An example of a vertical)))
9O Soil Mechanics,\" Basic Conceptsand Engineering Applications

(a)

(d)

Figure 3,17. Methods to control the flow rate and quick condition.

impermeable blanket is a sheet pile system driven into the soil, either at the toe or at another
position within the base of the dam (Figure 3.17(c)). piles are more effectivein increas-
Sheet
ing tim factor of safety but less effective in decreasing the flow rate. A toe drain shown in Fig-
ure 3.17(d) may also be used to significantly increase the factor of safety but it will result in a

higher flow rate that may or may not be desirablefor the project. A combination of the above
methods may be used to optimise the length of the base which is designed mainly to provide

stability against hydrostatic and gravity forces.

Example 3.12
ResolveExample 3.11 with a horizontal impermeable blanket of 8 m length added to the sys-
tem at the upstream side.
Solution:

5.0m 5.0rn
I I
\177 \177-I
i \177-1\177 I

\1775.0m

8.0 m
; \177\"---,\177--.___L ; '\177_---k--\"\177 \\

impermeable stratum
Figure 3.18. Example 3.12.)))
 The Movement of Water through Soil 91

The transformed length of the horizontal blanket is: 8.0 \303\227

0.625 = 5.0 m.
Figure 3.18 shows
the flow net drawn schematically for the transformed section. The number of equipotential
lines were kept the same: N\177\302\242
= 8, and
Nf= 4.
Using Equation 3.27: 1.6x10 -5 x5.0x1000x3600x24x4/8
q= = 3456 I/day.
The flow -
rate is reduced by: (4320 3456) / 4320 = =
0.2 20%.

Exatnple 3.13

For the concrete dam shown in Figure 3.19 construct the flow net under the dam, calculatethe
flow rate (in m3/day) and the distribution of uplift pressure on the base of the dam. The base is
1 m below the original ground surface and a sheet pile 7 m long is driven into the soil at the
-5
upstream side. h = 9 m, k = 4 x 10 m/s, 'h,,,
= 21 kN/m 3.

Solution:

The flow net is shown in Figure 3.19, from which: Nd = 10 and Nf

= 4.
Using Equation 3.27:
_5
q = 4.0x 10 x9.0x3600x24x4/10 m3/day (per = 12.44
run). metre

The total head across the length of the base is formulated as follows:
h = 10,0 + 1.0(depth of foundation) -
0.9ha, where nd is the number of equal drops of 9.0/ 10
= 0.9 m up to the point of interest on the base. The distance of the intersection points of equi-
potential lines from the left corner of the dam and the relevant heads and water pressures are
presented in the table below. For point (1) we may use interpolation or, conservatively, simply
accept the higher magnitude corresponding to nd= 3.

Poim i 2 3 4 5 6 7

x (m) 0.00 4.75 9.75 15.00 I9.25 22.50 25.00

h (m) 8.3 7.4 6.5 5\1776 4.7 3.8 2.9
u (kPa) 81.4 72,6 63.8 54\1779 46.1 37\1773 28.5

\"\177 100

\177
50
\177 0
0 5 \1770 15 20 2,5 30

25.0 m
\342\200\242

\"tO.O m.

20.0 m

I \177
\"
I I

Figure 3.I9. Example 3,13.)))

92 SoilMechanics:BasicConcepts and Engineering Applications

The diagram of uplift pressure is also shown in Figure 3.19. The average hydraulic gradient
between the base and ground surface on the downstream side is:
i = Zkhfl = 0.9fl.0 = 0.9, ic = Y'fYw =
(21.0-9.81)f9.81 = 1.14,

Factor of safety against the quick condition = 1.14/ 0.9= 1.27.

3.5 MATHEMATICS OF THE FLOW IN SOIL

3.5.1 Continuity equation for a steady]low

Figure 3.20shows a three dimensional element of soil in which the total head of water is h.

Assuming a steady state flow, the quantity of water entering the element paralleI to one of the
(e.g. x-axis) is: qx of velocity in the x direction and Ax
= Vx Ax, where
axes Vx is the component
= & is the sectional area of the element
dy \303\227 perpendicuIar the direction of flow. Hence:

Oh Oh
v Oh = = v = = -k (3.30)
x=kxi x=-k x Vy kyiy -ky z k:.i z z
Ox' --\177y, -\177z

The negative sign is arbitrary and causes the value of h to decrease in the direction of the ve-

locity. Substituting Vx into the flow rate equation we obtain:

qx
= -kx ohdv\177dz (3.31)
Ox

The amount of flow that exits the element in the x direction is qx + dqx in which dqx is the

variation of the flow rate in the x direction and is expressed by:

OZh
= -k dxdv \177dz (3.32)
dqx x 2
Ox

Assuming a zero volume change the flow rate entering the element is equal to the flow rate ex-
iting the element. Therefore,the continuity condition can be written as:

qx + qy
+
qz =qx +dqx +qy +dqy +qz +dqz
dqx
+ + dqz = Avatume = 0 (3.33)
dqy

qy+dqy
z
qz+.dq7

qx qx+dqx

Figure 3.20. Steady flow through a three dimensional element.)))

 The Movement of Water through Soil 93

Substituting Equation 3.32 into the continuity condition we have:

32h OZh \" O2h
= 0
k
x + + kz (3.34)
\177 ky 2 \177
Ox OY OZ'\177

In two-dimensional flow, where flow in, say, the y direction is zero, the corresponding term in
the continuity equation vanishes. Equation 3.34 therefore becomes:

02h OZh
--
OZh
__
kx32h+k,-S-\177-=O,
2 \177
or--+
2 \177
=0 (forkx=kz) (3.35)
3x 3z- Ox Oz
The solution to this differential equation yields the total head within the seepage zone. It is
seen that the distribution of the total head is independent of the coefficient of permeability
(when kx
=
kz) and depends entirely on the geometry of the problem. The continuity condition
expressed by Equation 3.35 may be arranged in terms of the components of the velocity by
substituting Equations 3.30, resulting in:

Ov---Lx + Ovz = 0 (3.36)

Ox

Equation (or 3.36) is the classical

3.35 differential equation of flow in two dimensions and is
referred to as the
Laplace equation. Traditional methods of solving this equation include direct
integration and numerical analysis. Direct integration is carried out only for seepage problems
with simple boundary conditions by using complex fu\177xctions. Whilst there are an infinite
number of solutions to the Laplace equation, the boundary conditions control the selection of
the function. For one dimensional flow, in the x direction:

02h -
2
0 (3,37)
Ox

Equation 3.37 may be integrated twice to give the following linear relationship for the total
head: h = Ctx + C2, where C1 and C2 are integration constants found by substituting two
known boundary conditions into the linear equation. These known boundary conditions may
be selected as the total heads hi and h2 corresponding to two sections with xt = 0 and x2 = L
where L is the distance between the two sections. Substituting these values into the linear
equation of total head we get:
Ct =-(h t-h2)/L, C 2=h l

The total head between two sections ofxl and x2 is represented by:
h =
-x(h I
- h IL
2) + hl (3.38)

By differentiating Equation 3.38 and applying Darcy's law, the velocity is found to be:

0h - \177h
vx = = -k = k h\177 h\177 - k
kxi x x x x (3.39)
\177 L L
This shows that the velocity is constant between sections 1 and 2 and is consistent with the ba-
sic assumption stated previously (Figure 3.i).)))
94 Soil Mechanics: Basic Concepts and Engineering Applications

3.5.2 Potential and stream functions

Equations 3.30 represent the projections of the velocity dimensional system and im-
in a three
ply that the vetocity is perpendicular to the constant through the point.
h passing
The analyti-
cal sotution of two dimensional flow through an isotropic materiaI expressed by Equation 3.35
is obtained by the introduction of two functions qb and W to satisfy Equations 3.30.

\177h
- O@ -
v x = -k . v. = -k --

where \177 and W are called potemialfimctiot\177 and streamfunctton respectively. Substituting
Equations 3.40 into 3.35 we obtain:

z +--Z-3-= 0 f3.41)
Ox Oz-

The solution of Equation 3.41 will yield a result for qb(x. z), which can be expressedas follows
based on the definition of
@(x.z) = -kh(x, z) + C (3.42)
where C is a constant depending on the boundary For different but constant
conditions. values
of qb. such as qb\177, qbz, and qb.3, Equation 3.42 represents a series of curves for which the po\177en-
tialfunction and total head are constant i.e. equipotemial lines. From the definition of the
function q-' it is seen that it is tangential to the velocity and represents the direction \177f the flow
and flow lines. By comparing Equations 3.40 and 3.42, we obtain:

-- = - (3.43)

These are called the Cauchy-Riemann equations whosesolution yields a series of stream func-
tions or flow lines for constant values of\177, W2 .... assigned f\177r \177(x, z.). Taking the derivative
of the first term \177f Equation 3.43 wi\177h respect re) z and the second term with respect to x and
addin\177 the results, then:

-- + =0 (3.44)
\342\200\242
3a-2 \177

:\177- x Stream or Flow lines

Figure 3.21. Calculauon of the quantity of flow between two flo,.,, lines.)))
 The Movement of Water through Soil 95

which means that the stream functions are in the form of the Laplace equation. Figure 3.21
shows two flow lines representedwith \1771 and W2- At an arbitrary point within these flow lines
the velocity v is shown with its projections in the xz plane. With the arrangement shown for
the positive variation of x and z along the equipotential line, one of the projections of the ve-
locity is in the negative direction of the corresponding plane axis. The flow dq corresponding
to the length ofds on the equipotential line is:

dq = vxds = v x xdz - v. xdx

Substituting velocities from Equations 3.40 we obtain:

dq = \177--\177-W
dx + 3---\177-\177 =
dz d\177d, or
3x 3z

Aq = f\1772 d\177 =
\1772
-
\177l (3.45)
1

This shows that the flow rate between two successive flow lines is equal to the difference be-
tween the stream functions. If the stream function is known in terms ofx and z then the veloc-
ity components can be calculated using Equation 3.40. Conversely, with known functions for
the velocity components, the stream function can be obtained by:
\"3u? \"oq\177
W = dx + dz +C (3.46)
J-\177--x J\177
A useful result is obtained by substituting Equations 3.43 into Equation 3.41:

= 0 (3.47)
3z3x Ox3z
This indicates that a stream function obtained from Equation 3.46 automatically satisfies the
continuity equation. Equations 3.35 (or 3.41) and 3.44 are of differential
a set equations for
which closed form solutions are availableonly in very solutions are simple cases, Analytical
obtained by using the theory of complex variables, For example,for an impermeable dam with
a base length of b, and a total head loss of h, resting on a permeable layer, similar to Figure
3.12 but with infinite depth, Wu (1966) reported the following closed form solutions for po-
tential and stream functions:

x2 a_2 - x2 z
2
1, + = 1 (3.48)
2 2 2
(b (b cosh 7tq\") (b sinh 7\177W)2
cos-\177) (bsin-\177) h h

In general, the solution of two-dimensional flow is achieved by numerical techniques such as

finite element and finite difference methods. In the finite element method the seepage zone is
divided into discrete elements with common nodes. The parameter of interest (h, qb or \177) is as-
sumed to vary within each element according to a linear or a non-linear function. A solution is
obtained for specified boundary conditions that covers the entire seepage zone. In the finite
difference method the seepage zone is divided into a network of nodes usually known as grid

points. The partial derivatives at each point are replacedwith a finite difference approximation)))
96 SoilMechanics:BasicConcepts and Engineering Applications

b.q
\"'\"\177,.\177, Flow line
Equipotential line
\177,/ \177 12 !..\177._._

Figure 3.22. Deflection of flow lines at the boundary between two soils with different permeabilities.

that results in a linear

algebraic equation. With known boundary conditions, and application of
the continuity at the boundaries with
condition unknown numerical values for the specified
function (h, @ or a
g), unique solution can be obtained for grid point values. The variation of
or gr is assumed
h, \342\200\242 to be linear betweengrid points, which is a major disadvantages com-
pared to the finite element method. However,the finite difference method is easy to formulate
and a useful alternative to the analytical solution.

3.5.3 The transfer conditions

The line ab in Figure 3.22 represents the boundary between two soils with different k values.
The flow lines at the kl side are deflectedas they k2 side. Thisis represented
enter into the by
different values of \177, the angle between a flow line and the direction normal to the boundary.
The flow net at the kt side is comprised of squares of length l\177. At the k2 side the squares

change rectangles
into t2 by 13 which represent the distance between the equipotential lines and
flow lines respectively. Considerthe case where one flow line and one equipotentiat line inter-
sect at a point B located on the boundary. According to the continuity condition the flow rate
inside a flow lane at the kt side is equal to the flow rate in the deflected lane at the k2 side.
= Aki = = = Aki = (t
Aql (t\177xl)k\177AtHt\177 ktk\177h, zXq2 3 xl)k2Ah/t 2
From the geometry of Figure 3.22, 12 and t3 can be expressedin terms of 131, [32and tl:

12
=
BCsinl32 = l 1 sinl32 lsin[31, 13
=
ABcos[32 = l t cos[32/cos[31
Substituting 12 and 13 in the flow equations and equating kql by zXq2 we obtain:

kt / tan = k / tan 132---+ k 1 tan = tan

13\177 2 132 k2 fl t (3.49)

3.5,4 Numerical analysis of two-dimensionat flow using the finite difference method
In this method two-dimensional flow, represented by the continuity Equations 3.35 or 3.41 or
the differential equation of stream function in Equation 3.44, is approximated by finite differ-)))
 The Movement of Water through Soil 97

ence equations. this section

In we deveIop numerical vaIues for the soIution of the Laplace
equation for head given in Equation
total 3.35. Let \177x and Az be the change in the coordinate
of a point represented by i andj as shown in Figure 3.23(a). Restricting the Taylor's series to a
finite number of terms, the total head at points (i + l,j) and (i
-
1,j) representing +\177x and
respectively, are:

Oh zS'x2 02h 3\"x3 03h \177 O4h

=
h\177+\177,\177h\177,\177
+ + (3.50)
+\177(\177)\177,\177 +\177(\177)\177,\177 3\177(0x\177)\177,\177

Oh \1772 \1772h \1773 \1773h \1774 04h

= - +
h\177_w h\177,\177
+ (3.51)
\177(\177)\177j \177(0x\177)\177,\177 -\177(\177)\177,\177

Adding Equations 3.50 and 3.51, we obtain:

02h -- hi+l,j +hi-l,j-2hi, j (3.52)

(\177)i,j (\177)2

Similarly, for flow in the z di\177ction, we have:

( 02h) = hi'j+l +hi'j-1-2hi'j (3.53)

2 i,j \177
OZ (\177z)

Substituting \177uations 3.52 and 3.53 into the continuity \177uation 3.35:

hi+l,J +hi-l,j -2hi,j hi.j+l +hi,j-i -2h i ' j

\177k = 0 (3.54)
kx z
(\177)\177 (\177)2

For convenience we assume \177 = \177, and kx = kz and therefore:

i,j+l

\177Zi-l,j i, i ---]-- -]--\" i-l,i Ii, j i+l,j

i=+l,j ; -
\342\200\242 i
|

\177 x (b)

(a)

AX

'sz
i.--1, l ]i,f

(d) (e)

Figure 3.23. The basic finite difference grid.)))

98 Soil Mechanics: Basic Concepts and Engineering Applications

= hi+l'J
+
hi-l'J + hi'j+l + hi'j-I
h.i, j
4
(3.55)
This equation shows that, in a square grid, the total head at every grid point is the average of
the total heads at the four adjacent grid points. It also implies that the total head at every point
in the grid represents the average total head within a flow lane of thicknes.s
\177x or \177z, and that
the point is located on the centre line of the flow lane (Figure 3.23(b)). This means that if we
apply Darcy's law in the form q = Aki, and the continuity condition in the form 5q = 0 (at
every point) we can obtain Equation 3.55. Note that in this case, the hydraulic gradient is the
first derivative of the total head expressed in finite difference form.
Equation 3.55 can be modified (using the concept of flow lane and continuity condition) for
the case when the point located on an
(i,j) is impermeable boundary (Figure 3.23(c)):
+
hi-I'J
+ 2hi'j+l
= hi+l'J (3.56)
hi, j 4

lf the point (i, j) is locatedat the base of a vertical impermeable boundary with negligible
thickness (e.g. sheet pile), as illustrated in Figure 3.23(d), the total head at the grid point (i,j+
1) can be taken as the average of the total heads at the left and right of the boundary:

- hi+t.j +hi-l,j +O.5[(hi, j+l)L + (hi,j+l)R]+hi, j_ 1

hi, j (3.57)
4
If the point (i,j) is located on the boundary of two soils with different permeabilities, and the
boundary is parallel to the horizontal axis, then Equation 3.55 becomes:

= + + 2kB
+
2kT
%(fzi+l.j hi-l.j
hi,j fzi \")+l ,-1)
hi \"\177 (3.58)
kr +k\177
kv +k B

where are coefficients of permeability

kT and kB at the top zone and the bottom zone of the
boundary respectively (Figure 3.23(e)). For each unknown head at an interior or a boundary
grid point, there is a condition expressed by Equations 3.55 to 3.58. Thus, the total number of
linear equations and unknowns are equal, and a unique solution can be obtained. Traditionally,
the set of equations was solved using a relaxation method, which included the readjustment of
assumed values of the unknowns to satisfy the appropriate condition at each grid point. How-

ever, with the improvement of computational tools, a direct solution is now preferred.

Example 3.14

Figure 3.24(a) shows a sectionof a concrete dam, with an 8 m base, resting on a permeable

layer of thickness 8 m. Taking the impelaneable stratum as the datum for total heads, calculate
the total heads at the grid points using the finite difference method and construct the equipo-
tential line for a head loss of 1.25 m. Repeat the solution using the relaxation method by a re-
fined \177id (occupying the same area) with bx = Az = 2 m.

Solution:
A mesh ofAx = Az = 4 m is constructedto evaluate the total heads at the grid points. Since the
flow net in the permeable layer is symmetric, only one half of the section is considered.)))
 The Movement of Water through Soil 99

2,3 \177
3,3 ,4,3 ,53.00 113.o0 ,13.00 0.50
\177
4.0 I
I
I
I
I !

.... r2,2....
\177 \177
4,2 , B \"1 0.50
r ....
3,2
8.0 m
].1_2._61_ _ \177
!
1\" \177 \177_3._00_
, -' 11.96\177 \177
---\177
t l I
I
I I I I I i I I
I I I I
\177t,1 \1772,1 \1773,1 ,4,1 113.00 ',12.49 1.73 10.50

impermeable stratum
(b)
(a)
Figure 3.24, Example 3,14.

conditions: = = = 5.0 + 8.0 = 13.0m. For a symmetrical flow net:

Boundary hi, 3 h2,3 h3. 3

h4, I
=
h4, 2
=
h4.3
= 2.5 + 8.0 - 10.5m.
Assume at the left boundary (grid points 1,1-1,2 and 1,3) the flow rate is zero and the pore
= =
at these hi. 3 13.0 m.
are =
pressures points hydrostatic: hi. 1 hi. 2
Applying Equation 3.55 at the interior grid points 2,2 and 3,2 and Equation 3.56 at the grid
points 2,1 and 3,i on the impermeable boundary, and substituting the known boundary heads:

-h2.1 + 4h2.2
- = 26.0, - + = 23.5
h3. 2 -h2, 2 h3, I 4h3, 2
- - = 13.0, - =
4h2,1 2h2, 2 h3,1 -h2,I +
4h3,1 2h3,2 10.5
Solving for the unknowns: h2,l = m, h3,1 = 11.73 m, h3.2 = J. 1.96 m.
12.49m, h2.2
= 12.61

Thc total hcad for thc equipotential line with a 1.25 m loss is 13.0 - 1.25= t 1.75 m. The
dashed line in Figure 3.24(b) shows the equipotential line for h - 1 t .75 m, where the locations
of points A, B, C and D have been established by linear interpolation between the two grid
points at both sides of the point of interest. For the finer grid the first estimations of the total
head at the grid points are shown in Figure 3.25 (first row). The first iteration is applied using
Equation 3.55 for the interior points and Equation 3.56 for the points on the impermeable
boundary. These results are presented in the second row whilst the third row shows the results
of the fourth and final iteration. The equipotential fine corresponding to h = 11.75 m is repre-
sentedby the solid line, while the dashed line represents the previous analysis with the coarse

grid. Sample calculations for the first iteration for grid points 3,2 (interior) and 3,1 (on the im-
permeable boundary) are:
/4 --) h4.2 + h2, - = 0.0
h3,3 + \177,1 4h3.2
=
h3, 2 (ha, 2 + h2.2 + h3, 3
+
h3.1) 2 +

But as the selected vaIues are not the exact values, a residual R is defined according to:)))
100 Soil Mechanics: Basic Concepts and Engineering Applications

2.0m

2,5 4,5 5,5 6,5 7,5

1,5 3,5
13.00 13.00 13.00 13.00 13.00 ] 11.62 10.50

2,0 m 11.62
I 11.58
\302\261
[,4 2,4 3,4 4,4 5,4
/ ;6,4 7,4
13.00 12.90 12.80 12.64 12.2g / 11.49 10.50
12.90 12.79 12.60 12.2k\177 t\" 11.41
12.90 12.78 12.59
/
11.39
12.ff
1,3 2,3 3,3 4,3 7,3
,,'
13.00 12.8O 12.61 12.30 11.23 10.50
12.79 12.61 12.35 I1.97 11.28
12.82 12.61 12.32
1,2 2,2 3,2 4,2 r5\177 6,2 7,2
13.00 12.75 12.55 12.19 t1.84 11.17 1O.50
12.74 12.5't 12.18 \1771.76 11.17
12.77 12.51 12.18 11.74 11.16

1,1 2,1 3,1 4,1 5,1 6,1 7,1

3.00 12.70 12.49 12.12 11.73 11.11 0.50

12.75 12.48 12.15 11.73 11.13
12.75 12.47 12.14 11.70 I1,13
Figure 3.25. Example 3.14.

R3, 2
=
h4, 2 +
-
h2, 2 + h3, 3 + h3.1 4h3, 2 12.19 + 12.75+ i2.61+ 12.49
= - 4x i2.55 = -0.16m.
The corrected value of h3.2 to satisfy the continuity condition is:
= = 12.55+ (-0.16)/4 = 12.51m. For
h3, 2 h3, 2 + R3, 2/4 grid point 3,1:
= + +
h3, I (h4, ! I h2,
2h3,2)/4,
= + + \177-
12.12+12.7 - 2x12.55 -4x12.49 = -0.04 m,
R3, t h4, t h2,t 2h3, 2 -4h3,1
= = 12.49 + (-0.04)/ 4 = 12.48
h3,t h3,t + R3,1
/ 4 m.

More precise values may be obtained if the left boundary is moved further to the left in order
to justify the assumptions made on the total heads for the points on this boundary.

3.6 SEEPAGE THROUGHEARTH DAMS

3.6.1 Entry and exit conditions

An idealized section of a typical earth dam is shown in Figure 3.26. It is constructed on an im-
permeable base and the seepage zone is limited to the embankment soil only. The height of the
water at the upstream side is h whilst at the downstream side, the water table is assumed to be)))
 The Movement of Water through Soil 101

Impermeable base Free draining

Figure 3.26. An idealized section of a permeable earth dam.

on the ground surface. A toe drain may be constructed at the downstream side to control the

uppermost flow line or phreatic surface. length Its is designed to stop the phreatic surface
from exiting at the downstream slope. The main objective is to establish the phreadc surface
and the flow net within the dam in order to calculate the flow rate as well as the parameters
necessary for a stability calculation, such as pore pressure and seepage pressure. Since the co-
ordinate system shown in Figure 3.26 does not agree with the conventions of the previous sec-
tion, the relevant equations are modified for this arrangement. The problem involves three
cases of transfer conditions:
1. Entry of water from the k = ,,\177zone (water) to the soil with coefficient of permeability of k.
2. Possibleexit of water from the soil to the air (k = ,,',) at the downstream slope.
3. Entry of water from the soiI to the free draining at toe with k = \177.
The entry and exit conditions for a boundary between a soil of permeability of k and a zone of
infinite permeability may be established by constructing a flow net around the boundary sixni-
lar to Figure 3.22. However, in this section, a completeproof of the transfer conditions is not

presented. Figure 3.27(a) shows a case in which water enters from the upstream face to the
soil. As the upstream face is an equipotential line, the flow line must start perpendicular to this
face. The flow line rises up a little above the normal line to the face as shown in the figure. If

there is a gravel wall between the water and the earth dam, the phreatic surface will be tangent
to the water level in the gravel and wilt drop downwards, as shown in Figure 3.27(b). Note
that the downstream side of the gravel wall does not represent an equipotentiat line. Figure
3.28(a) shows the entry condition from the soil to the air at the downstream face of the earth

\342\200\242 \"
w line \177 FI ne i
\"\1770 70:
soil
\177 Coarse gravel
,\177ct
< 90
\302\260
__
/ X > 90? i
i

(a) (b))))

Figure 3.27, Entry conditions from the water to the soil.

102 Soil Mechanics: Basic Concepts and Engineering Applications

_\17790\302\260,, _
Flow line\177\"\177
/ l-ree
t!i
.,/ I Flow line

\177g \177 \177\177\177

a< 90 \177

Free draining
(a) (b)

\177
. i i Flow line Vertical/
\177
Flow \342\200\242
Free i
I
i
/\302\260il
S-\"\177.\177
Soi\177 draining i ; V
--\177-
i ; ....90 / Free draining
7

(c) (d)
Fibre 3.28.Ent\177 conditions from the soil \177othe a\177rand drain.

dam where the phreatic surface is tangent to the slope. Figures 3.28(b) to 3.26(d) show the exit
conditions from the soiI to a drain with a horizontal, a vertical and an inclined surface respec-
tively. In all cases the flow line is tangent to a vertical line irrespective of the slope of the sur-
face of the drain.

3.6.2 The equation of the phreatic surface: the basic parabola

A traditional (Casagrande) method defines the phreatic surface as a parabola with its focus lo-
cated at the origin O of the xz coordinate system (Figure 3.29). This basicparabola can be de-
fined mathematically if the coordinate of one point on the boundary (or within the seepage
zone) is known, and is shown as curve FMCin Figure 3.29. Experimental investigati\1773ns have

Z
E F A/\177,\177..-\177
Directrix

Impermeablestratum Free draining

Figure 3.29. Casagrande method establish

\177o the phreatic surface in an earth dam.)))
 The Movement of Water through Soil 103

Dire ;trix

0 x
Impermeable stratum
p/2\177) \177--
(a) p
(b)'\177 \177---

Figure 3.30. Correction of the phreatic surface to satisfy the exit conditions.

shown that the intersection

parabola point of
and the water surface, point F, is lo-
the basic
cated such that FA = that the coordinates of point
0.3 EA, which F are known. The pa-
means
rabola has to be corrected at point A to meet the requirements of the entry conditions. The dis-
tance of any point M on the basic parabolafrom the focus is equal to the distance of this point
from the directrix, which is located at an unknown distance ofp from the z-axis.

MO =MN --> \177x2 + z

2
=x+p --,'.x = (z 2 -p2)/2p (3.59)
By substituting the coordinates of point F in Equation 3.59, the value ofp so obtained is:

P=\177F + Z2F -x,= (3,60)

The flow rate is estimated by constructing the flow net schematically as explained earlier,An
alternative solution is to assume a constant hydraulic gradient in the vertical sections:
q = Aki =
(zxl)kdz/dx, from Equation 3.59 dz/dx = p/z and:

q =kp (3.6\177)

The horizontal length of thetoe drain L must be sufficient to allow the basic parabola to be lo-
cated inside the earth dam. It is calculated by intersecting the equation of the basic parabola
with the equation of the downstream face z = tanct (x + L) and seeking the condition in which
both functions become tangent to each other:

L = p(1+ cot
2
00/2 (3.62)
In the absence of a toe drain the basic parabola intersects the downstream face at point B as
shown in Figure 3.30(a). In reality, the phreatic surface must be tangent to the downstream
face at point T with a distance a from the origin O. In the Casagrande method, the correction
length of zXa (Figure 3.30(b)) is found from Figure 3.31 which is based on experimental re-

sults. Note that the distance OB= Aa + a can be easily established,as the equations of the ba-
sic parabolaand the downstream face are both known.)))
104 Soil Mechanics: Basic Conceptsand Engineering Applications

0.4
0.35
0.3
0.25
0.2
0.15
0,1
0.05
30 60 90 120 150 180

\177(degrees)

Figure 3.31. Casagrande method to correct the phreatic surface.

Table 3.2. &\177/ (,Sa + a) in terms of aagle or.

(degrees) 30 60 90 120 t 50 180

\177al(\177Xa+a) 0.36 0.32 0.26 0.18 0.10 0

Table 3.2 shows the approximate values of &z / (&z + a) for given angles of et based on
Casagrade's experimental work. An estimation of An /
(Aa + a) can be made by assuming the
following linear relationship of the form reported by Linsley et al. (1992).

z\177a/(2\177a + a) = (! 80 - \177z)/ 400 (3.63)

which is shown as the dashed line in Figure 3.3i. The top flow Iine can now be corrected by
sketching a smooth curve tangent to the basic parabola at/T and
at T(Figure downstream face
3.30(b)). An alternative called the Dupuit
solution, method, defines the correction curve as fol-
lows: (a) at any vertical section the hydraulic gradient defined by dz / dx has a constant value,

(b) the curve passes through point F which is already defined in the construction of the basic
parabola,(c) The curve is tangent to the downstream face. It can be shown that the first condi-
tion yields the correction curve as a parabola in the form: x = Cl Z2 + C 2.
The integration constants CI and C2 are found by substituting the second and third conditions
in the above equation. Hence,the equation of the flow line adjacent to the downstream face is:

x = cos\177(z 2
-z2f)/2asin2\177+xf (3.64)

The coordinates of point T are x = a cosa and z = a sinct. Substituting these values into Equa-
tion 3.64 two answers are obtained, the positive value of which is the distance a.

a - (3.65)
cosc\177
2
ct sin ct)))
 The Movement of Water through Soil 105

O.Om 5

Impermeable stratum 60.0 m

(a)

(b)

Figure 3.32. Example 3.15.

The flow rate can be obtained by calculating dz / dx from Equation 3.64 and substituting this
result and zr = a sins into the flow equation:

q= Aki = (z
T xl)k(dz/dx) -\177 q
= kasin \177tan c\177 (3.66)

Example 3.15

A vertical section of a large earth dam is shown in Figure 3.32(a), Estimate the flow rate by
sketching the flow net and use the DupuJt assumptzon, kx =
Solution:
tan c\177\177-
60.0/[(200.0
- 20.0) / 21 = 0.666--\177o\177= 33.69 \302\260
. to Figure 3.29:
Referring
82.50 m,x \177-= 200.0 - 82.5 + 0.3\303\22782.5 - 60.0 = 82.25 m, zF = 55.0 m.
= 55.0/tan33.69 \302\260
=
EA

From Equation 3.60: p

= \"\17782.252 + 55.02 -82.25 = 16.69 m.
2
The equation basic parabolais: x =(z2 -p2)/2p
of the =(z -278.71)/33.38.
Figure 3.32(b) shows the flow net within the earth dam in which Nf = 3 and Na = 10.
q = khNf /N d = 3.0\303\22710
-6
x 24x 3600x55.0x 3/10 = 4.28m3/day.

the Dupuit condition applies: q = kp = 3 x 10 16.69 x 24 x 3600 = 4.33m3/day.

_6
When \303\227

Example 3.16

For the earth dam section shown in Figure 3.33 calculate the values of a and zla and the flow
3 -6
rateacrossthedaminm/day, kx=kz=3xlO m/s.

Solution:)))
 106 Soil Mechanics: Basic Conceptsattd Engineering Applications

14.0 m

160.0 m Impermeablestratum

Figure 3.33. Example 3.16.

tan ot = 32.0/[(160.0 - 14.0)/ 2] = 0.43836 --\177c\177= 23.67 \302\260.

Referring to Figure 3.29:
m, x F = 160.0- 68.44+ 0.3 \303\227
\302\260 = 112.09 m,
EA = 30.0/tan 23.67 = 68.44 68.44 ZF
- 30.0 m.
From Equation 3.60: p
= \177/112.092 + 30.02 - ! 12.09- 3.94m.
The equation
=
of the basic parabola is:x (z - p2)/2p
2
= (z 2 -3.942)/7.89.
The length OB Aa = + a is
calculated by substituting:
= OB cos 23.67 \302\260
and z\177 = OB sin23.67
\302\260
into the equation of the basic this results
xt\177 parabola;
in OB = 46.90 m. Using Equation 3.65:

- 112.091cos23.67\302\260
- \177/1i2.092/cos 2 \302\260 2 \302\260
_
a 23.67 -30.02/sin 23.67 25.46 m,
&a =46.90 - 25.46 = 21.44 m.
q = ka sin ct tan ct = 3.0x 10.6 x25.46
xsin 23.67
\302\260
x tan 23.67
\302\260
x 3600x 24 = 1.t6 m3/day.

3.6.3 Design considerations to control the.flow rate

Different types of earth dams are shown in Figure 3.34 (Wilson & Marsal, 1979;Roberson, et

aI., 1997). The major factors involved in the control of the flow rate are the coefficient of per-
meabilities of the dam material and the base soil and the height of the water at the upstream
side. Horizontal or inclined drains, or a combination of both, can be used to control the down-
stream seepage, as shown in Figure 3.34(a). A traditional section includes an internal core
constructed from compacted clay material. This core reduces the seepage and has back drains
to remove water to the downstream side (Figure 3.34(b)). In the case of a permeable base, the
length of the flow line is increased by constructing impermeable blankets or by developing the
central core to the permeable base to form a cut-off trench (Figure 3,34(c)). Sometimes the
horizontal and vertical impermeable blankets are combined to make a cut-off trench, as shown
in Figure 3.34(d). Alternatively, the horizontal blanket may be projected upwards along the
slope at the upstream side (Figure 3.34(e)); aspbaltic cement is sometimes used for this pur-
pose (Linsley, et al., 1992).In recent years synthetic fibres have been used to provide drainage
as well as impermeable blankets. Synthetics have also been used as a reinforcement to provide
stability \177othe upstream and downstream slopes. If modifications such as these are employed,
particular care has to be taken in the evaluation of the overall factor of safety of the structure.
Any increase in the pore pressure will tend to reduce the factor of safety against shear failure.)))
 The Movement of Water through Soil 107

Impermeable dam

stratum Imperme,a,,bl\177-\177 ra}n

Impermeable blanket \177'/ \\\\ \177'\177
(a)
Central core Permeable base
[..J

Impermeable strat um

(d)
Impermeable stratum
n
(b) I mpe r me ab le\177'\177-\177-/\1777-\177D ra)
blan ket'\177'\177\"/// \177

Permeable base

Impermeable stratum

(e)
Impermeable stratum
(c)

Figure 3.34. Different types of earth dams.

3.7 PROBLEMS

3.1 In a constanthead permeability test, a cylindrical sample 100 rrun in diameter and 150

mm high is subjected to an upward flow of 540 ml/min. The head loss over the length of
the sample is 360 ram. Calculate the coefficient of permeability in ln/s.
=
-4
Answer: k 4.8x10 trds

are: diameter of = 20 ram, di-

3.2 In a laboratory falling head test, the recorded data the tube
ameter of the cell = 100 ram,
length of the sample
= 1000 ram. The head measured from

Flow
Overflow,,\177-\177

[::::\177,. \177 Discharge

\177
k
225m2\177Sof\177 % 150mm

\" \"

Figure 3.35. Problem 3.3.)))

108 Soil Mechanics.\" Basic Concepts and Engineering Applications

the top level of the sample dropped from 800 mm to 600 mm within 1 hour and the
of the water was 30\302\260C.Calculate the coeeficient of
2 permeability
at 20\302\260C.
temperature
-3 2 -3
rl
= 1.005 x 10 N.s/m (at 20\" C), rl
= 0.801 xl0 N.s/m (at 30\302\260C).
-6
Answer: k = 4.0x10 rrds

3.3 For the test arrangement shown in Figure 3.35, calculate the volume of water discharged
z
in 20 minutes. The cross sectional areaof the soil is 4000 mm and k = 4.0 mm/s.
Answer: Q = 23.04I
3.4 A long trench is excavated paralfel to a river, as shown in Figure 3.36. The soil profile
consistsof a permeable soil of thickness D confined between two impermeable layers.
Initially the water level in the trench is the same as that in the river. Water is pumped out
of the trench at a flow rate of q. (a) Formulate q in terms of the geometrical parameters
-5
shown in Figure 3.36, (b) for L = i00 m, D = 5 m, and k = 4 x 10 m/s, calculate the
flow rate corresponding to a drawdown of Dw = 2 m, (c) calculate q when the water table
in the trench is 2 m belowthe surface of the permeable layer.

Answers: q = kD(zw - D)I L (for full flow), q -- k[2D(z.\177,. - D) + D 2 - (zw - D w) 2]/2L

(for the water table in the trench under the surface of permeable layer), 0.346 and 0.622
3
m /day/metre run of trench.
3.5 A well of diameter 0,3 m is constructed to the full depth of an unconfined aquifer of
thickness of 150 m. The water table is 10 m below the ground surface. A pumping test of
12 m3/hour has resulted a drawdown of 10 m. Assuming ro = 400 m, calculate the coef-
ficient of permeability of the aquifer. If the flow rate increases to 18 m3/day, and in the
absence of any other data, what will be the best estimate for the drawdown in the well?

Answers: 3.1 x 10-6 m/s, 15.3 m

3.6 A pumping test carried out

m thick confined aquifer
in a 50
(well dia. = 0.6 m) results in
a flow rate of 600 I/rain. of the impermeable layer above the aquifer is 20
The thickness
m and the original water level in the well is 2 m below the ground surface (which is also
the top of the impermeable layer). Drawdowns in two observation wells located 50 m
and 100 an from the well are 3 and 1 m respectively. Calculate: (a) the coefficient of
permeability of the aquifer, (b) the drawdown in the well, (c) the radius of influence.

[-
\177
]\177

Figure 3.36. Problem 3.4.)))

 The Movement of Water through Soil 109

k = k = 5 \303\227
10
-5 rrgs
(a) x z
.5 m (b) k x = 5 x 10-5 m/s,
-5
k z = 3 x 10 m/s
_'\1776

(a) (b)
Figure 3.37. (a) Problem 3.8, (b) Problem 3.9.
-5
Answers: 1.1 x 10 m/s, 17.8m, 141,3 m.
3.7 A soil profile consists of three layers with the properties shown in the table below. Cal-
culatethe equivalent coefficients &permeability paraIlel and normal to the stratum.

Layer Thickness (m) kx (parallel, m/s) kz (normal. m/s)

-6 -6
1 3.0 2.0x10 1.0x 10
-8 -\177
2 4\303\2670 5.0x 10
-5
10
2.5\303\227
-5
3 3.0 3.0x 10 1.5xl0
10
-6 -8
Answers: kx = 9.6 x m/s, kz
= 6. [ [0
\303\227 m/s
3.8 For the pile system shown in
sheet Figure 3.37(a), calculate the flow rate in m3/day by
constructing the flow net in the two conditions (a) and (b) shown.
Answers: 9.82 m3/day, 7.60 m3/day

3.9 A concrete dam retains 8 m of water as shown in Figure 3.37(b). Calculate the flow rate
in re-/day by constructing the flow net under the dam. k = 5x 10 m/s.
Answer: 12.15 m3/day

3.0 1.0 m
kx= kz=6 x 10-7m/s
\177L \177

Impermeable stratum
100.0 m

Figure 3.38. Problems 3.10 & 3. \177

1.)))
110 Soil Mechanics: Basic Concepts and Engineering Applications

3.10 For the earth dam section shown in Figure 3.38, calculate the flow rate in m3/day.

Answer: 0.18 m3/day

3,11 For the earth dam sectionshown in Figure 3.38, calculate the minimum Iength of the toe
drain required to ensure that the phreatic surface becomes tangent to the downstream
face.

Answer: 8.65 m

3.8 REFERENCES

AS 2368.1990. Test pumping of water wells. NSW, Australia: Standard Association of Australia+
AS 1289.6.7.1.1999+ Methods soils for engineering
of testing purposes: Soil strength and consolidation

tests-Determination of permeability of a soil-constant head method for a remoutded specimen, NSW,

Australia: Standard Association of Australia.
AS 1289.6.7.2. 1999.Methods of testing soils for engbleering purposes: Soil strength and consolidation
tests-Determination of permeability of a soil-falling head method for a remoutded specimen, NSW,
Australia: Association of Australia.
Standard
AS 1289.6.7.3. 1999. Methods of testing soils for engineering purposes: Soil strength and consolidation
tests-Determination of permeability of a soil-constant head method using a flexible wall permeameter.
NSW, AustraIia: Standard Association of Australia,
ASTM D-5084+2000.Standard rest methods for measurement of hydraulic conductivity of saturated
porous materials using a flexible wall permeameter. West Conshohocken, PA: ASTM.
ASTM D-2434-68. 2000. Standard rest method for permeability of granular soils (constant head). West
Conshohocken, PA: ASTM,
ASTM D-4044, I996. Standard test method for (field procedure)for instantaneous change in head
(Slug): Tests for determining hydraulic properties of aquifers. West Conshohocken, PA: ASTM.
BS 5930.1981+ Code of practice for site investigation. London: British Standard Institution,
Head, K.H. 1986+ Manual of soil laboratory resting. London: Pentech Press.
Linsley, R.K+, Franzini, J.B., Freyberg, D.L. & Tchobanogtous, G. I992. Water-resources engineering.
New York: McGraw-Iiill.
Roberson, J.A., Cassidy. J.J. & Chaudhry, M.H. 1997, Hydraulic engineering. New york: John Wiley &
Sons,
Wilson, S.D. & Marsa] R.J. 1979.Curresnt trends h\177
design and construction of embankment dams. New
York: ASCE+
Wu, T.H. 1966, Soil mechanics. Boston: Allyn & Bacon.)))
CHAPTER 4

Shear Strength of Soils and Failure Criteria

4.1 INTRODUCTION

This chapter describes the shear strength characteristics of soils and related failure criteria.

Soits fail either in tension or in shear; however, in the majority of soit mechanics prob-
lems, only failure in shear requires consideration.The shear strength along any ptane is
mobilized by cohesion and by the angle of internal friction, collectively referred to as
shear strength parameters. If, at a point atong a specified plane, the shear stress becomes
equal to the peak shear strength along that plane, then the soil will fail at that point and
large shear strains will result. A part of the soil on one side of this plane, called the failure
plane, wifl slide relative to the other side, bringing about collapse of the soil structure. A

failure criterion is a mathematical relationship between the state of peak stress and the
shear strength parameters. When combined with the principles of solid mechanics, a fail-
ure criterion can be used in different types of stability problems to predict (collapse) loads
that cause the failure. The post peak strength behaviour of the soit may be predicted by us-
ing the critical state models that are now common for specific types of soils.

4.2 MOHR-COULOMB FAILURE CRITERION

4-2.1 The Mohr' circle of stress

Figure 4.1(a) shows the sign conventionfor two-dimensional stresses. From equilibrium
of the forces acting on blocks 2 (Figure 4.1(b)), the normal
1 or and shear stresses on any

plane P, with angle ot to the x-axis, are determined by the following equations:
-
\177= (o z + Ox)/2 +[(o z Ox)/2lcos2C\177 + \"rxz
sin 2or (4.1)

- 2\177: - 20:
\"c =
[(\177z \177x)
/ 2]sin
zxz cos (4.2)
Combining these equations to eliminate the parameter or, we obtain:

era
+ \177--Crx 9 2
(or
\177x
)2 + z2 = + z:,z
2 (_,_\177_)_
which is the equation of a circle in the o-'r coordinate system (Figure 4.1(c)).This circle is
called Mohr's circle of stresswhose centre and radius are defined by s and r respectively:

s= (ql +q3 )/2 = (o: +Cix}/2 (4.3)

111)))
 Soil Mechanics.\" Basic Concepts and Engineering Applications

-- Pot

(b)
0\"
z 1

Figure 4. I. Mohr's circleof stress.

/ cr -Crx \", 2
t = (4.4)
q(\177)-+
The major and minor principal stresses are given by:

131
= s +t, o 3 = s-t (4.5)
whilst the angle of the major principal plane to the x direction is definedby:

o=ltan- ]( 2zxz- ) (4.6)

2 \177r
z
O\"
x

The minor principal plane is definedby 13 + 90 \302\260.

Equations 4.1 and 4.2 show that the state
of stress in any specified direction o\177is found by knowing the normal and shear stresses in

two reference directions. If the crl and (Y3 stress directions are taken as referenceaxes
(Figure 4. [(d)), then:

=
\177Y c\302\260s2\177
(\177l +\1773)/2+[(\1771 -\1773)/21 (4.7)

= - sin 2ct
\"\177
[(\177Yl \177Y3)
/ 2] (4.8)

Example 4.1
An element of soil is subjected to the two-dimensional stresses shown in Figure 4.2(a).)))
 Shear Strength of Soilsand Failure Criteria 113

O'p
=79.8

= 31.6
\177,_..\177____\177Q
= -31.6
\177OP

TQPli x\177 I

\177PQ

\177p

= 50.0
z \177\177\303\227 = 79.8
20,0
op
I = 31,6
\177PQ
l\177xz\177

\1773=
43.0 \177 6z = 100.0
= 70.2 = -20.0
6Q Xzx
= -37.6
COO

(c) (d) (All in kPa)

Figure 4.2. Example 4. I.

Determine (a) the normal and shear stresseson the P and Q planes which are orthogonal,
(b) the magnitudes and directions of the major and minor principal stresses, (c) the corre-

sponding c\177values of P and Q if the principal axes are taken as the reference axes.

Solution:
(a) Using Equations 4.1 and 4.2:
=
c\177,\177 (100.0 + 50.0) - 50,0)
/ 2 + [(100.0 / 2] cos 120.0
\302\260
+ 20.0 si n 120.0
\302\260
= 79.8 kPa,
= -
z,,, [(100.0 50.0) / 2] sin 120.0 20.0 cos 120.0 31.6
\302\260
- \302\260
= kPa.
= (100.0 + 50.0)/ 2 + [(100.0- \302\260
50.0) / 2]cos 300.0 + 20.0 sin 300.0
\302\260
= 70,2 kPa,
OQ
= [(100.0-50.0)/ 2]sin 300.0 \302\260
- 20.0 cos 300.0\302\260
= -31.6 kPa.
zO
Figure 4.2(b) shows the normal and shear stresses on the PQ planes.
(b) From Equations 4.3 and 4.4:

s = (100.0 + 50.0)12 = 75.0kPa, t = J(100.0- 50.0)2 + 20.02 = 32.0kPa,

2

o I
= s +t = 75.0 + 32.0= 107.0kPa, o 3 = s -t = 75.0- 32.0= 43.0kPa.
The direction of major principal plane is:
-1
0 = l/2[tan (2'rxz
Io
z
- 50.0)]= 19.33
Ox) ] = 1/2[tan-1(2x20.01100.0- \302\260.

(c) In ch and o3 coordinatesystem the angle o, for the plane P is:

\302\260 \302\260
= -19.33 + 60.0 = 40.67 \302\260.
\177p

A summary of the results is shown in Figures 4.2(c) and 4.2(d).)))

114 Soil Mechanics: Basic Conceptsand El\177gineering Application,s

\177, Yield Ea ure

Hardening

\177--
(a) Shear Strain

Figure 4,3. Stress-strain behaviour of a soil element subjected to simple shear.

4.2.2 Stress-strainmodelsfor soils

Figure 4.3 shows an element subjected to a variable shear stressz and a constant
of soil
normal stress \177. The deformation of this element
behaviour is represented by a plot of
shear stressversus shear strain. At small shear stressesthe response of the element is lin-
ear until the yield point is reached; beyond this point plastic straining begins to occur
(work hardening). After the yield point the shear stress continues to increase and eventu-
ally reaches a peak value (or failure point), that is, the maximum shear strength of the
element corresponding to the applied normal stress. Further shear deformation causes the
shear stress to decease (work softening) to an ultimate value representing a critical state in
which the material fails while maintaining constant volume. Failure planes may develop
before or after the critical state leading to a constant shear strength called the residual
strength. However, this state is achievable only with very large strains. In overconsoli-
dated soils, failure planes may develop immediately after the peak point and the critical
state may not be obtained and the state of the element moves towards the residual. Nor-
really consolidated soil may not have a peak strength and the critical state may be obtained
at relatively low strains. Consider, now, a soil mass subjected to a system of external load-

ing which increases from an initial value until the mass fails. Every element of this mass
goes through the same deformation pattern shown in Figure 4.3; however, the shear

strength at a given point depends on the mag\302\242aitude of the load transmitted to this point.
Due to the complexity of the behaviour shown in Figure 4,3, some idealizations must be
made so that predictions of shear strength and the maximum value of the external loading
can be made. A few of the more common idealizationsare shown in Figure 4.4.

4.2.3 Mohr-Coutomb faiture eriterfo\177t

Coulomb (1776) suggested that the shear strength of a soil along the failure plane could
be described by:
- c+ crtan
z/. \177 (4,9)

where xf is the absolute value of the shear strength on the failure plane (corresponding to
the peak point in Figure 4,3(b)), a is the stress normal to the plane, c is the cohesion and 9
is the angle of internal friction of the soil (or internal friction angle). The two parameters e
and e are calledshearstreng#t parameters.)))
 Shear Strength of Soits and Failure Criteria 115

Rigid-plastic work
V
\177 softenin\177=g \177

(a) (c) (d)

Figure 4.4. Common models for soil behaviour.

To understand the concept behind Equation 4.9, consider the two blocks A and B (Figure
4.5) of unit area that are in contact with each other and subjected to the normal and shear
stresses shown. The contact surface is not smooth and contains frictional asperities. Under
a constant normal stress, the shear stress is increasedfrom zero to the maximum \"rj; forcing
the two blocks to slide along their contact area. When cr = 0, the shear stress has to be mo-
bilizedto a maximum value of c to make the sliding possible. If the friction angle between
blocks A and B is 0 then, for the values of o\" > 0, \"t has to be increased to overcome the re-
sistance to sliding \177\303\227 tan0 caused by friction. Consequently, the summation of c and \177\303\227

tan0 represents the maximum shear stressneeded to slide the two blocks on the plane of
contact. In a real soil, if a predetermined sliding surface is forced to occur, the soil below
and the soil above the failure plane will not act as rigid materials but will deform, causing
a volume change around the sliding surface. Moreover, the internal friction angle is not
the same as the friction angle between solid particles as it depends mostly on the inter-
locking mechanism between the solid particles. In a coordinate system with cr plotted

horizontally and x vertically, Equation 4.9 is represented by the line shown in Figure
4.6(a). This equation was originally written in terms of total stress and was only partially
successful in predicting the shear strength of real softs. Coulomb's failure criterion was
subsequently redefined as:

= c'+
\"rj.
or'tan o' (4.10)
\177
where \"rf is the absolute value of the shear strength, o is the effective normal stress, c\" is
the effective cohesion and 0\" is the effective angle of internal friction of the soil (Figure
4.6(b)). In the total and effective stress conditions,
both the shear stress is taken solely by
the particles since the liquid in the voids, which is normally water, has no resistance to
shear. The tensile strength of soil is commonly ignored and therefore cohesion is the
minimum shear strength at zero normal stress. Coulomb'scriterion does not apply in the
case of a normal tensile stress.

Figure 4.5. Mechanical concept of sliding.)))

i 16 Soil Mechanics.\" Basic Conceptsand Engineering Applications

(b)

Figure 4.6. Coulomb'sfailure criterion.

The shear strength parameters correspondingto the total and effective states have different
values. However, in the case of clean sand, it is reasonable to assume c\" = c = 0 and #' = 0.
\302\260 \302\260
The magnitude of @' or @ in a sandy soil varies from 25 to 45 depending on the soil type
\302\260
and density whereas, in a clay soil, it may vary from 0 to 25 depending on the soil type,
moisture content, and drainage conditions. In undrained conditions the cohesion (c,) var-
ies from 10 kPa to i00 kPa or more, depending on the density of the soil. Coulomb's fail-
ure criterion is also used to represent the residual strength and the subscript r is used with
each term in Equations 4.9 or 4.10to identify them as the shear strength parameters in this
state. Figure 4.7 shows the total and effective states of stress at the peak strength (failure
point) represented by Mohr's circles. It is apparent that the shear stress at every plane in
the total stress Mohr's circle is the same as in the effective stress Mohr's The dif- circle.
ference between the normal stresses in two perpendicular directions in the total and effec-
tive states are equal:

Thus, the radius of both the totalthe effective Mohr's circles are identical.
and The hori-
zontal distance of the two is equal to pore pressureu. Any point F on the failure
circles
envelope represents the normal and shear stresses on a failure plane at a specified point in
soil. These stresses must also satisfy the equilibrium conditions at the point, which is rep-
resented by Mohr's circle of stress. This implies that, at failure, Mohr's circle of stress
must be tangentto the line expressed by Equation 4.9 (or 4.10). This condition, referredto
as the Mohr-Coulomb failure criterion, is shown in Figure 4.7. The angle between the
failure plane F and the major principal plane is c\177=(90\302\260+q{)/2 =45\302\260+@'12.Due to

Figure ,t,7. Mohr-Coulomb failure criterion.)))

 Shear Strength of Soils and Failure Criteria 117

symmetry, there failure plane at angle -o: from the ff'l plane.
will be another From Figure
4.7, a relationship state of stress (0.'1
between and 0'3) or the
(0\"z, \177x and \"Cxz) and the
shear strength parameters c' and 0\" may be formulated by equating the radius of Mohr's
circle to the distance of the centre of the circle from the failure envelope:
t' = \177/[(0.r. -ox)/2l
2
+'tx2z =OBcosdo'+OCsindo'

Considering that - \302\242r

\302\242r
z x
= -0.
x, OC
=
(0.\177
+ 0.\177)/2 and OB = c' then:
0.\177

, , 2
\342\200\242
+ = [2c, cosqb,+ (\177,z + 0.x)sln ]
q\177) (4.11)
(0.\177 _o.1:)2 (2\177.xz)2

This equation can be expressedin terms of the principal stresses O-'land 0'3:

0.1 -0.3
= 2c'cos0' +(0.1+ 0.3)sin0\", or:

= tan
2 ,+ + 2c'
\302\260
+ 0'/2)
c\177 c\177 (45 \177,/2) tan(45 (4.13)
\302\242 \302\242 2
=
0\"3 0\"1
tan (45\302\260-0'/2) -2c'tan(45\302\260-dp'/2) (4.14)

Example 4.2
At a point in a soil mass, the total vertical and horizontal stresses are 240 kPa and 145 kPa
respectively whilst the pore pressure is 40 kPa. Shear stresseson the vertical and horizon-
tal planes passing through this point are zero. Calculate the maximum excess pore pres-
sure to cause the failure of the point. What is the magnitude of the shear strength on the

plane of failure? The effective shear strength parameters are: c\" = 10 kPa, 0' = 30\302\260-

Solution:

Check if the initial state of stress is below the failure envelope:

= 240.0 - 40.0 = 200.0kPa, \177145.0 - 40.0 = 105.0kPa. Refcrring to Figure 4.7:
\177 0.3
CF = OBcos qS' + OCsin qb'
= 10.0cos 30,0
\302\260
+ [(200.0 + 105.0) / 2]sin 30,0
\302\260
= 84.9 kPa,
t\" = radius of Mohr's circle = (200.0 - 105.0)2 / = 47.5 kPa < 84.9 kPa.
Hence, the initial state of stressis below the failure envelope.
0.1 = 240.0 kPa, 0.'t = 240.0
- 40.0 - Ue, 0\"1 = 200.0 - Ue

0.3
= 145.0 kPa, 0.'3= 145.0- 40.0- Ue, 0\"3 = 105.0 - ue

where Ue is the excess pore pressure.Substituting the majorand minor effective principal
stresses into Equation 4.13 and noting that c\177= 45
\302\260
+ 0' / 2 = 60\302\260:

200.0 - u =
(105.0 - u ) tan
2
60.0
\302\260
+ 2 x 10.0x tan 60.0
\302\260
--> u = 74.8 k_Pa.
e
e e

o\"
!
= 200.0-u
e
= 200.0-74.8 = 125.2kPa, 0.1
=
105\"0-Ue
= 105.0-74.8 = 30.2kPa.
Using Equations 4.7 and 4.8:
cy\" = (125.2+30.2)/2+[(125.2
- 30.2) / 2]cos 120\302\260
= 53.9 kPa,

= 53.9tan30\302\260+ 10.0 = 41.1 kPa, or: \177'y =[(125.2 -30.2)/2]sin120.0

\302\260
=41.1 kPa.
zf
Example 4.3

A cubic sample of clay soil is subjected to the plane strain loading of: 0\"z = 0,x = 80 kPa,
Zxz(initial)
= 0. The shear stress \"Cxz is gradually increased untiI the sample fails. If the
shear strength parameters are c' = 40 kPa and 0' = tS\302\260,
determine (a) the maximum shear)))
118 Soil Mechanics: Basic Concepts and Engineering Applications

stress applied on the four faces of the sample, (b) the shear stress on the failure plane, (c)
the magnitudes of the major and minor principal stresses, (d) the minimum value of 0,z
= \177'x to avoid any tensile stresses in the sample and the corresponding magnitude of \"txz.

Solution:

(a) At the initial state when '\177xz

= 0, Mohr's circle of stressis represented by a point with

=
\302\242fzo'x
= 80 kPa on the horizontal axis. Using Equation 4.11:
2 = 2
(2\"r,\177) [2 x 40.0cos 15.0\302\260
+ (80.0 + 80.0) sin 15.0\302\260] -\177 \"\177zx
= 59.3 kPa.

(b) Using Equation 4.2 with the sign convention of Figure 4.1 (c):
= 195.0 \302\260, = -59.3cos195.0 \302\260
=
2o\177 =90.0\302\260+2(45.0\302\260+ 15.0\302\260/2) 57.3 kPa.
\"\177f
(c) (r'l = 80.0+ 59.3
- 139.3 kPa, 0\"3= 80.0 - 59.3 = 20.7kPa.
(d) From Equation 4.13 with 0\"3 = O:
0\"1
= 2 x 40.0 tan (45 \302\260
+ 15\302\260/2)= 104.2 kPa, 0\"\177= 0\"\177
=
( 104.2 + 0.0) / 2 = 52.1kPa.
The radius of Mohr's circle = \177xz= (104.2 -
0.0) / 2 = 52.1kPa.
Example 4.4

An element of soil 8 m below the ground surface is adjacentto a long vertical retaining
wall where the water table is 2 m below the ground surface. The retaining wall is sub-
jected to a horizontal displacement towards the outside of the wall until the soil behind the
wall fails. Assuming that shear stresses cannot develop on the surface of the retaining
walI, determine: (a) the total pressure appliedto the retaining walI through the element, (b)
the direction of the failure plane, (c) the shear strength on the faiIure plane passing
through the point. Properties of the soil are:
= 1.8 Mg/m 3, 9sat
= 2.0 Mg/m3, c' = 10kPa, and = 30 \302\260.
9dry \177'

Solution:

(a) There is no shear stress on the vertical plane, thus there will be no shear stress on the

horizontal plane passing through point. the The mechanism of the failure indicates that the
vertical plane is the minor principal plane:
\1771
= (2.0 x 1.8 + 6.0 \303\227 x 9.81=153.0
2.0) kPa, u = 6.0 x 1.0 x 9.81 = 58.9 kPa,

0,1 = 153.0
- 58.9 = 94.1kPa. Substituting cr'\177 in Equation 4.14:
0,3 = 94. ltan2(45.0
\302\260
- 30
\302\260
/ 2) - 2x 10.0x tan (45.0
\302\260
- 30.0 \302\260
/ 2) = 19.8 kPa,
\1773
= total
pressure to the retaining wall through the element = 19.8+ 58.9 = 78.7 kPa.
\302\260
(b) 2c\177= 2(45.0 + 30.0\302\260/2) = i20.0 c\177=
\302\260, 60.0 This
\302\260. means that the soil behind the retain-
\302\260
ing wall could slide downwards at an angle of 90.0 - 60.0 \302\260
= 30.0 \302\260
to the wall.
\302\260
(c) Substituting 2c\177= 120 in Equations 4.7 and 4.8:
=
\302\242y\"
(94.1 + [(94.1- i9.8)/ 21cos
+ 19.8)/2 120.0 \302\260
= 38.4
kPa,
= 38.4tan30\302\260 = [(94.1 -
19.8)/2]sin 120.0 = 32.2
\302\260
+ 10.0 = 32.2 kPa, or: kPa.
'rj. '\177f

4.3 LABORATORY SHEAR STRENGTH TESTS

4.3.1 Direct shear, simple shear and residual shear tests

in the direct shear test points failure envelope are determined

on the directly by applying a
constant normal stress to the sample and measuring the shear strength on the predeter-
mined failure plane. A diagrammatic section of a direct shear test apparatus, called the)))
 Shear Strength of Soils and Failure Criteria 119

shear box (or split shear box), is shown in Figure 4.8(a). This box is made from metal and
has two halves that can move horizontally relative to each other. In this way the soiI
specimen inside the box is sheared in a horizontal direction. The \177lormal Ioad N is applied
vertically to the loading platen by means of a hanger with dead weights. Horizontal dis-
is
placement applied by a motorized system causing shear stress on the horizontal shear
plane. Thin metal plates, either grooved or solid with or without porous stones, are placed
at the top and the bottom of the specimen to ensure a uniform distribution of shear stress
and to control the drainage conditions. In plan, the box is square with a side of 60 ram, 64
mm or 100 mm whilst its total thickness of 2(} nun is shared between the upper and the
lower boxes. Circular sections are also manufactured with diameters the same as the
square boxes. Soils with Iarger size aggregates are tested in a shear box apparatus that can
accommodate specimens up to 300 mm square. These larger shear boxes have recently
been recognized by some standards (e.g. BS 1377). In a traditional shear box apparatus the

vertical and horizontal loads and displacements are measured mechanically; recent im-
provements include electromechanical measurement using an electronic recording system
that accepts analogue inputs. Figure 4.8(b) shows the mechanical concept of a simple
shear apparatus, referredto as a research shear apparatus (Powrie, 1997), whose advanced
design include the electromechanical measurements of forces and displacements and in-
cludes a pore pressureport. Square or circular shear boxes are also used to obtain the re-
sidual strength that is achievable only with very large horizontal displacements. Such
large displacementsare provided by reversing the direction of the shear displacement and

/
Normal load N
.L Toothed
I
J,, .. plates
Load measuring \177 J/
device \\ [-\177'
........
\177-\177r\"
force
Shear
T

' \"\177
Rolle\177 \177\"

Figure 4.8. (a) Diagrarrmaatic section of a shear box, (b) simple shear mechanism, (c) ring shear
mechanism.)))
120 Soil Mechanics: Basic Concepts and Engineering Applications

repeating the forward and reverse displacements until the residual strength is constant.
Based on the mechanism shown in Figure 4.8(c), a ring residual shear apparatus has been
developed that accommodates 5 mm thick ring samples with inner and outer diameters of
70 mm and 100 The sample is compressedvertically
mm respectively. between porous
bronze loading platens by means of a counterbalance system providing the normal stress.
A rotation is applied to the base plate to shear the sample, forming a shear surface close to

the upper platen that is artificially roughened to prevent slip at the platen-soil interface.
Soil specimens are either undisturbed or remoulded, depending on the type of the mate-
rial. Undisturbed specimens are prepared from fine-grained soils. However, in a coarse-

grained soil, the test sample is remoulded and particular care is taken to achieve similar

physical properties to that of the undisturbed soil, In order to construct the failure enve-

lope, a test has to be repeated several times, each with different normal stress and, ideally,
identical samples. Measurementsrecordedin a direct shear test include shear force, hori-
zontal relative displacement between the two split halves (or shear displacement)and ver-
tical displacement of the sample. The stress data is presented in (\177 or o\,") \"t coordinate sys-
tem whilst the volume change is represented by volumetric strain, vertical displacement of
force N, void ratio, or specific volume. Specific volume v is a dimensionless parameter
and represents a volume in which the solids occupy a unit volume. From a phase diagram
it can beshown that v = 1 + e. Any of the volume-related parameters may be plotted
against shear displacement to evaluate volume change properties. The volumetric strain,
void ratio or specific volume cannot be calculated accurately as the distribution of volume
change within the thickness of the sample is not uniform but concentrated mostly above
and below the failure plane.
The application of the direct shear test in the assessment of the strength properties of
soils in plane strain conditions has been recommendedby many standard codes, symposia
and textbooks, e.g. the Geotechnique symposium of 1987 drew special attention to this
application (Jewell, t989). Disadvantages of the conventional shear box have been high-

lighted by early researchers including Terzaghi & Peck (1948) and Roscoe (1953). These
include the effects of progressive shear failure initiated from the ends, the vnknown stress
distribution within the box and the creation of tensile zones under light normal loads. Ar-
thur & Aysen (1977) showed that the volume change is concentrated above and below the
failure plane within a zone approximately 6 mm thick for a medium size sand. Results ob-
tained from X-ray radiography showed that failure is progressive and starts from the ends
and develops towards the centre of the sample. A non-uniform distribution of shear stress
has been identified as one source of inaccuracy in the direct shear test and results in an
underestimate of the angle of internal friction. The author believes that the major reason
for the non-uniform stress distribution is due to the lack of friction at the ends, which
causes the failure plane to divert from the predetermined horizontal plane and creates a re-
taining wall failure pattern in the sample. A detailed analysis of the direct shear test can be
found in Arthur et al. (1977 and 1988). Jewell (1989) recommended that the internal fric-
tion angle measured in a split shear box should be specified as the direct shear angle rather
than the plane strain shear angle. However, this method of analysis does not include the
effects of deposition direction, which changes the resulting internal friction angle within a
\302\260
range of 6 in medium size sand (Arthur & Aysen, 1977; Wong & Arthur, 1985).

Example 4.5

From the following direct shear test data determine: (a) the internal friction angles at peak)))
 Shear Strength of Soits and Failure Criteria 121

and residual states, and (b) the principal stresses and their directions inside the box at the

peak strength when the normaI stress is 100 kPa. The shear box is 64 mm square in plan.

Vertical load (kg) Shear force (N) at peak Shear force (N) at residual

50 399,8 228.7
100 801.9 457.4
I50 1214.0 686.1

Solution: (a) Calculate normal and shear stresses and tabulate the results:

Normal stress (kPa) Shearstrength (kPa) at peak Shear strengti\177 (kPa) at residual

119.7 97.6 55.8

239.5 195.8 1 l 1.7
359.2 296.4 167.5

The results are plotted in Figure 4.9, from which \177'

= 39.25 \302\260
and \177'r
= 25.00 \302\260.
\302\260
(b) 20t = (90.0 + 39.25 =
\302\260)129.25 \302\260.
Using Equations 4.7 and 4.8:
100.0= (o\177 +
\342\200\242
- cr \" / o
o3)/2+[(o \177 3) 2]cos 129.25 ,
\302\260
= -
100.0\303\227 tan 39.25 /
[(\302\242r\177 5 ) 2] sin 129.25
0- \302\260.

Solving for 0-'! and o-'\177: 0.'1 = 272.2 kPa, o\"3

= 61.2 kPa, o. = 129.25\302\260
/ 2 = 64.62 \302\260.

4.3.2 Triaxial compression test

Figure 4.10 shows a schematic view of a triaxial celI and a cyIindrical soil sample. The
soil sample is protected by a thin rubber membrane and is under pressure from water that
occupies the volume of the cell. This confining pressure enforces a condition of equality

on two of the total principal stresses i.e. 0-2 = 0.3.Vertical stress is applied via a loading
ram, and therefore, the total major principal stress 6t is the sum of the confining pressure
and the deviator stress applied through the ram. In a traditionaI triaxial compression test,
the confining pressure 63 is kept constant whilst the major principaI stress 0-t is increased
incrementally by loading ram until the sample fails, The loading ram has a deformation-
control mechanism in which the horizontal surface of the sample is subjected to a con-
trolled rate of deformation; the associated force in the ram is measured by means of prov-
ing rings or Ioadceils.The rate of applied deformation is specified by the relevant code to

4O0

300

\177\"

\177 200

IO0

0
\177
\177 ! !

0 100 200 300 400)))

o\" (kPa)

Figure 4.9. Example 4.5.

122 Soil Mechanics: Basic Concepts and Engineering Applications

Loading ram
Air release

Perspex ..........\177l[r\177 il sample

Rubber

Drainage II

measuremen\177

Figure 4,10. Schematic view of a triaxial cell.

to avoid generation of any pore pressure in drained tests. Facilitiesto measure the pore
pressure and volume change at any stage of the test are available.To ef-
eliminate any end
fects on the results, of a specimen is twice its diameter.
the height Specimen diameters are
normally either 38 or 100 rnm, however some cells have been manufactured to accommo-
date larger diameter specimens,Specimensare either undisturbed or remoulded depending
on the type of the material. To construct the failure envelope for a soil, a test has to be per-
formed several times with different confining pressures using, ideally, identical samples.
Recordedmeasurements include deviator stress at different stages of the test, particularly
at failure (peak deviator stress)and the critical state (ultimate deviator stress),vertical dis-

placement of the ram, volume change and pore pressure, Stress data is presented in ((7 or
o\-\177t") coordinate system, as shown in Figure 4.11, and for each test, a Mohr's circle of ef-
fective stresses at the peak is drawn. The common tangent to the circles, which is obtained
graphically or by simple mathematics, represents the failure envelope.
The stress data may be presented in a t' = (\302\242r'l
-
or'3) / 2 and s' = (\1771 + \177'3) / 2 coordi-
nate system (or in their equivalent total stresses), as shown in Figure 4.12. This is most
convenient ira stresspath method is used. In this presentation the failure envelope passes

Failure envelope

t2'

\177f'3 0/1
\177
Figure 4.1 I Presentation of stress results from traditional triaxial test.)))
 Shear Strength of Soils and Failure Criteria 123

\177
t

Failure envelope

a'

o\177
3
\177

Figure 4. ]2. Presentation of effective stresses in z'- s' coordina[e system.

through the maximum shear stress point (A) of each circle. The parameters 0' and c' can
be related to the equivalent shear strength parameters of or' and a' according to:
0' = sin-1(tan or'), c\" = a' / cos q\177\" (4.15)

An alternative presentation of the stress state for both the peak and critical states is in the
q'
= o\"i - or'3 and effective mean stress p\"
= (o\"1 + 2o\"3) / 3 coordinate system (or in their

equivalent total stresses). From the definitions ofp' and q\" we obtain:

= =
\177 p'-q'/3, \177r\177 p'+ 2q'/3 (4.16)

Substituting 0\"3 and o'1 into the Mohr-Coulomb failure criterion (Equation 4.12) we ob-
tain the following linear relationship between q' and p':

q'= p'[6sinO\" /(3-sin#')]+ 6ccos#' /(3-sin#' ) (4.17)

Equation 4.17 represents the Mohr-Coulomb failure criterion in the q', p' coordinatesys-
tem. The stress-strain model and volume changeproperties are investigated by plotting the

deviator stress and volume

change against the axial strain. It is sometimes convenient to
normalize the stressthe deviator stress by or'3, or q' by p'. if the cohesion
data by dividing
(either effective or total) is zero then the stress-strain curves for different cenfining pres-
sures all fall close to a single curve. From the Equation 4.13 we can see that, in the ab-
\302\260
sence of any cohesion, the ratio of o\", to constant
o\"3 is and is equal to tan2(45 + q\177\"
/ 2).
The cross sectional area of the sample, which varies during a test, is given by:

=
A
Ao(I-\177v)/(I-\177I) (4.18)

where A0 is the initial cross sectional area, 8v = AV / V0 (AV being volume change and V0

being the initial volume) is the volumetric strain and \1771

= AL / L0 (&L being the axial de-
formation and L0 being initial height) is the axial strain. In undrained conditions, the vol-
ume change is zero, and therefore Equation 4.18 becomes:

A= A - (4.19)
0/(1 \1771)

4.3.3 Common O,pesof triaxiat test

Different types of triaxial tests have been developed in order to provide a comprehensive
description of the behaviour of the soil under different drainage conditions and loading
rates; these are:)))
124 Soil Mecltanics: Basic Concepts and Engineering Applications

Drained test: There is no condition on the degree of saturation and drainage alloweddur-
ing an experiment. Incremental loads are applied slowly to avoid building up excessive
pore pressures.
Consolidated-Undrained test: In this test, the soil sample is fully saturated and is consoli-
dated under a specified constant total stress during which drainage is permitted (to facili-
tate the consolidation). The applied pressure normally represents the in-situ conditions.
After completion of consolidation the sample is shearedin undrained conditions.

Unconsolidated-Undrained test: Here, the sample is fully saturated and no drainage per-
mitted during the test. Thus, the average moisture content remains constant and no volume

change occurs during the experiment. In consolidated-undrained and unconsolidated-

undrained tests a backpressure system may be used to increase the initial pore pressure to

insure full saturation. The magnitude of the backpressure does not affect the testresults.

Example 4.6
A consolidated-undrained triaxial test gave the following data (columns I to 4): Diameter
of the sample
= 38 ram, height of the sample = 76 ram. Pore pressures at peak points
are I0.0, 61.6, 113.2 kPa for or3
= 100, 200, 300 kPa
respectively.Determine: (a) the de-
viator stress-axial strain curve and Modulus of Elasticity of the soil, and (b) the shear

strength parameters (effective and total).

Axial de- Applied axial force (N) Deviator stress (kPa)

Axial
formation 6-3=100 6-3=200 6-3=300 o.3=100 63=200 o'3=300
strain
(ram) (kPa) (kPa) (kPa) (kPa) (kPa) (kPa)
0.00 0.0 0.0 0.0 0.0000 0.0 0.0 0.0
0.95 t26.3 155.7 183,7 0.0125 110.0 135.6 160.0
1.90 218.9 304.1 353.2 0.0250 t88.2 261.4 303.6
2.85 277.7 415.4 532.1 0.0375 235.7 352.5 451.6
3.80 308.8 475.8 622.4 0.0500 258.7 398.6 521.3
4.75 328.9 499.0 661.2 0.0625 271.9 412.5 546.6
5.70 332.3 512.9 689.3 0.0750 271,0 ,$18.3 562.2
6.65 335.6 518.6 701.9 0.0875 270.0 417.3 564.7
7.60 338.! 523,0 708.7 0.1000 268.3 415.0 562.4
8.55 341.3 528.9 715.7 0.1125 267.1 413.9 560.1
9.50 344.2 534.6 724.0 0. t250 265,6 4t2.5 558.6

600
\"\177
500
d.\177

\"\177\"400

\177 300

\177 200

\"\177 100

0,05 0.1 0.15

Axial strain,

Figure 4,13. Example 4.6.)))

 Shear Strength of Soils and Failure Criteria 125

Solution:

(a) Deviator stresses are calculated the corrected

from sectional areas and tabulated (col-
umns 6 to 8 in the above table). Sample calculation for axial deformation of 4.75,nurse:
-6 -\177
el = \177 //4 = 4.75 / 76.0 = 0.0625,Ao = (\177 x 38.02 ] 4) 10
\303\227 = 1.134115 x 10 m 2.

1.134115 / (1 - 0.0625)= 1.20972310

-3
A = x m 2.
At = 100 kPa deviator stress = (328.9/t000)l 1.209723x 10
-3 = 271.9kPa,
\1773 \177] -\1773
at \1773 -- 200 kPa - = 412.5 kava and at cr3 = 300 kPa - = 546.6 kPa.
similarly \177 \177 \177i cr\177

Figure 4.13 shows the variation of deviator stress against axial strain.
At \1773
= 100 kPa the initial Es is the best tangent to the curve from
ModuIus of Elasticity
the origin or is approximately equal to: Es = / 0.0125 (second 110.0
row) = 8800 kPa.
At or3 = 200 kPa, Es = 10848kPa, and at \1773 -- 300 kPa, Es = 12800kPa.

(b) CaIculate cr't an.d \1773 at failure for each test:

Test 1:\1773 = 100.0 - 10.0 = 90.0kPa, o\"1 = 100 + 271.9 - kPa. 10.0= 361.9
Test 2:\1773 = 200.0 - 61.6 = kPa, \177t
= 200 + 138.4
418.3 - 61.6 = kPa. 556.7
Test = 300.0 -
3:0\"3 kPa,! 13.2=186.8
\177'l
= 300 + 564.7 kPa.
- 113.2= 751.5
Mohr's of stress are shown
circles in Figure 4.14 from which: = 37 \302\260.
c\" = 0 and \302\242'

\1771 values at failure for tests 1, 2, and 3 respectively: crl= 10f3 + 271.9 -- 371.9kPa, 200 +
418.3 = 618.3 kPa, 300 + 864.7 kPa. 564.7=
Mohr's circIes of stress are shown in Figure

4.14 from which: c = 40 kPa, and o = 25 \302\260.

It\" the effective stresses are presented in the t'
and s\" coordinate then, from 4.14, we = 0 and o( = 31\302\260
system, Figure obtain a\"

Example 4.7

The following results were obtained from consolidated-undrained tests on specimensof a

saturated clay. Determine the shear strength parameters (effective and total).

\1773(kPa) ff'l-o\"\177(kPa) atpeak u (kPa) atpeak

I00 137 28
200 210 86
300 283 147

Solution:Calculate an.d
\177TY\177l 0,3 for each test:
Test 1:\177 = 100.0 + 137.0 - 28.0= 209.0kPa, \1773
= 100.0 - 28.0 = 72.0kPa.
Test 2: \177
= 200.0 + 210.0 - 86.0= 324.0kPa, O'z3
= 200.0 - 86.0 = 114.0k.Pa.

400

300

2OO

100 \", \",

0
100 200 300 400 500 600 700 800 900

\177,0\" (kPa)

Figure 4.14. Example 4.6.)))

126 Soil Mechanics: Basic Concepts and Engineering Applications

Test 3:\177'1
= 300.0 + 283.0 - 147.0= 436.0kPa, c(3 = 300.0 - 147.0= 153.0
kPa.

These results represent three Mohr's circles in the \177'-'\177coordinate system. The graphicaI
method may be replaced by the following simplified mathematicaI analysis:
Using the Mohr-Couiomb failure criterion of Equation 4.13:

(1) 209.0 = 72.0 \302\260

+ \177' / 2) + 2c\" /
tan(45\302\260+ \177' 2)
tan2(\17745
(2) 324.0 )
114.0tan'2(45\302\260+\177'/2)+2c'tan(45\302\260+\177'/2
\302\260 \302\260
(3) 436.0 t53.0 tan (45 + 0\" / 2) + 2c' tan(45 + 0' / 2)
This producesthree equations with two unknown values of c' and \177'. Solving (1) and (2),
(1) and (3), and (2) and (3) for c' and \177', and taking the average, we obtain c' = 1.6 kPa =
0.0 and \177'
= 28.3 \302\260.

For the total shear strength parameters:

Testl: \177I
= 100.0 + 137.0 = 237.0 kPa, \1773
= 100.0 kPa.
Test2:\17771= 200.0 +210.0 = 410.0kPa, =
\17773 200.0 kPa.
Test3:\1771---300.0 + 283.0= 583.0kPa, or3
= 300.0 kPa.
Substituting the test results into the Mohr-Coulomb failure criterion of Equation 4.13 and
solving for c and \177, we obtain c = 24.3 kPa and =
q\177 15.5 \302\260.

Example 4.8

The effective shear strength parameters of a fully saturated soil are c\" = 20 kPa and 0\"
= 25 \302\260.
In an unconsolidated and undrained triaxiai test, the cell pressure was 200 kPa and
deviator stress at faiIure was 107 kPa, Compute the pore pressure at failure.
Solution:
\177J'l
= 200.0 - u + 107.0 = 307.0 - u, 0\"3 = 200,0 - u, substituting into Equation 4.13:
307.0 -u =(200.0-u)tan2(45.0\302\260+25.0\302\260/2)+2x20.0\303\227tan(45.0\302\260+25.0\302\260/2),

u = 169.8 kPa,

4.3.4 Unconfined compressiontest

Unconfined compression is a specia] case of triaxiai compression where cr3 = 0. However,
rather than using a triaxial ceil, the test is carried out in a mechanical
specially apparatus
manufactured for this purpose. Soil samples are cylindrical with a diameter of 38 mm and
a height of 76 nun. This test is suitable onIy for fully saturated clays where Ou is zero and
the undrained cohesion can be found by constructing only one Mohr's circle. Unconfined
compressive strength (UCS, \1771 in Figure 4.15) is frequently used in stabilization studies to

Failure envelope

f -'\177 \177,=unconfined
streng,h

\177 \177pressive

Figure 4.15. Unconfined compression test.)))

 Shear Strength of Soils and Failure Criteria 127

ascertain the proportion of stabilizerrequired to satisfy a specific strength criterion. Sam-

ples used in this case are 100 mm in diameter and 200 rran high.

4.3.5 Stress path method

The presentation of stress data in either the t'-(s'or s) or the q'-(p'or p) coordinate systems
facilitates the monitoring of the stress increments and pore pressure variation during a tri-
axial test, which is a considerableadvantage. Consider a triaxial test specimen of normally
consolidated clay (point C in Figure 4.16(a)) that is consolidated under a confining
pres-
sure 0\"3. The confining pressure is greater than the preconsolidation pressure (maximum
effective stress under which consolidatedin the
the soil has been field) to ensure that the

soil will behave as a normally consolidated soil. At point C, s' = s = \1773 and the pore pres-
sure is zero. The progressof the undrained triaxial test is shown by path CE for the effec-
tive stresses and path CT for the total stresses. Both paths terminate on the corresponding
failure envelopes at points E and T. These points are the topmost points of the effective
and total Mohr' circles respectively.As a result, the angle of line CT with the horizontal is
always In a drained
45 \302\260. test the stress path coincides with CT but terminates on the effec-

tive failure envelope. In the q'-(p'or p) coordinate system the slope of the effective stress
path (in drained conditions) or total stress path (in undrained conditions) is 3 vertical to 1

horizontal, as given by:

Slope= q'/(P'-0\"'3) = (0.{- 0.;)/[(\177J\177 + 2o;)/3-0.;] = 3

The horizontal distance between CE and CT representsthe pore pressure according to:
s-s'= (0\"1
+ 0.3)/2-(0.\177 + 0.\177)/2
= +u+cr
(0\"\177 3 +u)/2-(0\" 1 +0.;)/2 =u

Figure 4.16(b) shows a typical stress path for an overconsolidated clay subjected to a con-
solidated-undrained triaxial test.

Example 4.9
The results of consolidated-undrained triaxial tests on three identical samples of clay are

tabulated below. Determine: (a) the effective and total shear strength parameters using the
t'-(s; s) coordinatesystem, (b) the stress path for both effective and total stresses.
Solution:

(effective)k/ envelope (total)

Failure envelope
.\177/ \177
(t\302\260tal)

\177
S \177,$ \177/,\177, \177__ s
s,,
C C

(a) (b)
Figure 4. I6. Stress path in consolidated-undrained test for: (a) normally consolidated clay, (b) over-
consolidated clay.)))
128 Soil Mechanics: Basic Concepts and Engineering Applications

Axial er3 = 150 kPa er3

= 300 kPa er3= 450 kPa
\177
straill t\" u S\" t\" u s t\" u $'

0,00 0.0 0,0 150.0 0.0 0.0 300.0 0.0 0.0 450.0
0,02 50.0 45.0 155.0 85.0 95,0 290.0 142,0 140.0 452.0
0,04 84.0 47.6 186.4 135,0 110,0325.0276,0 160.0 566.0
0,06 124,0 35,8 238,2 164,0 110.0354.0 382.0 144.0 688.0
0,08 153,0 22,0 281.0234,0 96,0 438.0 430.0 115.0 765.0
0.0972 440.6 101.4789.2
0.I0 176.0 12.0 314.0 271.0 81.0 490.0 440.0 101.0789.0
0.I2 192.02.0 340.0 296.0 66.0 530.0 440.0 90.0 800.0
0,I4 200,0 -6.0 356.0 314.8 52,0 562.8 434,0 80.0 804.0
0.i512 319.2 46,2 573.0 -
0.i545 204.6 -12,6 367.2 -

0,16 204,0 -13,0 367.0 318,0 42,0 576.0 426,0 72.0 804.0
0.I8 202.0 -17.0 369.0 308.0 31.3 576.7 416.0 66.0 800.0
0.20 195.0 -20.0 365.0 292,0 23,6 568.4 406,0 56.0 800.0
0,22 188.0 -23,0 361.0 279.0 17.5 561.5 400,0 52.0 798.0
0.24 184.0 -26.0 360.0 266.0 14.0 552.0 396.6 50.0 796.6
0.26 180.0-28.0 358.0 - 394.0 48.0 796.0
0.28 179.0 -30.0 359.0 - 391,0 47.0 794.0
0,30 178.0 -32.0 360.0- 388,0 47.0 791.0
0+32 177.0 -32.5 359.5 - - 384,0 46.5 787.5
0,34 174.0 -32.6 356.6 - - 380.0 46.5 783.5
0.36 171 -32.7 353.7 - - 378.0 46.2 781.8

The recorded data includes deviator stresses (twice of t') and pore pressures at specified

strains, except at peak points where the strains have been caIculatedfrom the recorded ax-
ial deformation. The values of s\" = (cr't + cr'3) t 2 are then computed and tabulated,
Check s\" at peak point for \1773 = 150 kPa: t' = (\177'1 - 0\"3) / 2 = 204.6 kPa, u = - 12.6 kPa,

s = (or1 + \1773) / 2 = (\1771 \1773+ 2\1773) / 2 = t' + \1773 = 204.6 + 150 = 354.6kPa,
-

s\" = s
- u = 354.6 - = 367.2kPa,
(-12.6)
Plotting the points in the s) coordinate system (Fig 4.17(a))we have:
t'-(s',
Total Stresses: \177= 23.8% a =
48.5 kPa. Substituting these values into Equation 4.15:
c =
= 26.2 \302\260,
\177 54.0 kPa. Effective Stresses:\177t\"= 29.1 \302\260,
a\" = 0.0, from Equation 4,15:

\177'
= 33,8 \302\260,
c\" = 0.0. The stress paths are shown in Figure 4.17(b).

500
400
300
200
100
0
200 400 600 800 1000 200 400 600 800 1000)))

(e'\177
+ e'3) / 2, (el + \1773) / 2 (kPa) 07'1 + \1777'3)
+ \17773)/ 2
/ 2, (\1777\177 (kPa)

(a) (b)
Figure 4.17. Example 4.9.
 Shear Strength of Soils and Failure Criteria 129

4.3.6 Pore pressurecoefficients A and B

Pore pressure coefficientsrelate the applied load to the excess pore pressure in undrained

loading. In an axisymmetric stress system where the loading is controlled by variation of

\177rl and o3, two coefficients are sufficient Pore pressurecoef-
to construct the relationship.
ficients are computed from the triaxial test resuIts and may be applied to a field situation

to predict excess pore pressures created by undrained loading. Figure 4.18(a) shows an

eIement of elastic soil under total principal stresses of ol, o2, and o3, with an initial pore
pressure of ui. AI1 three principal stresses are increasedequally by zX\177y,causing an increase
of Au in pore pressure (Figure 4.18(b)). The increase in effective stress in all three direc-
tions is: A\177 = AO - Au. In an elastic element subjected to these equal stress increments in
the three directions, the principal strains are defined by:

El
=
\1772
=
E3
= (1 -- 2g) (4.20)
where Es is the Modulus of Elasticity of the soil and t-t is Poisson's ratio. The volumetric
strain and voIume change are given by:
AV 3Ac( 3Act\"
E V =\177=EI+E2+E3 = \177 (1 -2p.) -\177AV = V
(1-2#) (4.21)
V E
s
The volume change in the pore water due to excess pore pressureAu is obtained from the
fundamental definition of volumetric strain for pore water:
A Vw
= A._\177_u
--7 AV = V A_.ff_u
w w (4.22)
Vw K K
where Vw and the volume and the buIk modulus
K are of the pore water respectively. In
fuji saturation, the volume of pore water is equal to the total volume of the element V mul-

tiplied by the porosity n. Therefore:

&u
= nV
\177,V,,, (4.23)
K
In undrained conditions no drainage is aIlowed, and consequently, the voIume change of
the soil skeleton and pore water must be equal.By equating Equations 4.21 and 4.23 we
can expressthe excess pore pressure Au in terms of the total incremental stress

\302\2601 \1771
+ A<3
\1772..+ AO\" O', + Ao 0-2
o2

\1773 \302\2603

\1771 \1771
+ A\177 0-1 +

(a) (b)
Figure 4.18. (a) Soil element, (b) excesspore pressure due to equal increase in \177, 0\"2, and cr 3,

(c) excess pore pressure due to increase in oh.)))

L
130 Soil Mechanics: Basic Concepts and Engineering Applications

i
Au = A\177 = BAcr
(4.24)
i+ nEs
3K(I- 2tx)

where B is termed the pressure

pore coefficient. Figure 4.18(c) shows an elastic soil ele-
ment in which only the major principal stress r\177l is subjected to an increase of Ar\177. Follow-
ing the development described above, we obtain:

1
Au
- = --1 B50.
A0. (4.25)
nEs 3
311-\177 - ]
3K(I 2\177t)

Considering the test path in a traditional triaxial compression test we can combine the two
cases to yield:

= + (A(\177 1 -
Au B[AO\" 3 A0.3)/3 ] (4.26)
-
where AO'3 represents the change in the confining pressure and hcrl ho'3 is the change in
the deviator stress. However,it is customary in soil mechanics practice to employ Equa-
tion 4.26 for the determination of the excess pore pressure,where both pore pressure coef-
ficients A and B are determined from data recorded during a triaxial test:

Au = B[A0.3 + A(A\177 I
-503) ] (4.27)

The parameters A and B are generally different from their theoretical values as the soil is
not a perfectly elastic material and the response depends on the level of the applied
stresses. The magnitude of parameter B depends on the degree of saturation, and since the
bulk of pore water
modulus is very high, B must theoretically be unity for saturated soils.

However, in partially saturated soils and in saturated soils with air bubbles in the pore wa-
ter, B decreasesto values less than unity. The magnitude of the parameter A depends on
the test path and stress level, and diverges from its theoretical value having positive values
for normally consolidated soils to negative values for overconsolidatedsoils.
Example 4. i0

Using the data of Example4.9, calculate the pore pressure coefficient A at every stage of
the three tests and show its variation with axial strain. Assume B = 1.
Solution:

A A
Strain {33=150 o3=300 0\"3=450 Strain 0.3=150 {33=300 0\"3=450
(kPa) (kPa) (kPa) (kPa) (kPa) (kPa)
0.00 0.000 0.000 0.000 0.16 -0.032 0.066 0.084
0.02 0.450 0.559 0.493 0.18 -0,042 0.051 0.079
0.04 0.283 0.407 0.290 0.2 -0.051 0\303\267040 0.069
0.06 0.1\1770, 0,335 0.188 0.22 -0.061 0.031 0.065
0\303\26708 0.072 0.205 0.134 0.24 -0.071 0.026 0.063

0,0972 - _0.115 0.26 --0.078 0.061

0.I0 0.034 0.149 0.115 0.28 -0.084 0.060
0.I2 0.005 0.111 0.103 0.3 -0.090 0.060
0.14 -0.015 0.082 0.092 0.32 -0.092 0.060
0.1512 - 0.072 0.34 -0.094 0.061
0.1545 -0.030 0.36 -0,096 0.061)))
 Shear Strength of Soils and Failure Criteria 131

0.6

0.5
0.4 = 300 kPa
\1773
\342\200\242
0.3 !\177
A

0 - \177 - kPa
-0.1 -
= 150 kPa
or3
-0.2 {

0 0.1 0.2 0.3 0,4

A\177al strain

Figure 4.19. Example 4. t0.

The results are tabuIated and shown in Figure 4.19.

Example4.11
Data obtained from a consoIidated-undrained triaxial test are: = 300
(\1773 klga, u (before con-
soIidation) = 276 kPa, u (at failure) = 108 kPa, (\177l - (\1773(at failure) = 475 k_Pa. Calculate
the pore pressure coefficients A and B.

Solution:

Using Equation 4.27 at the consolidation stage: 276.0 = B [300.0 + A (0)], B = 0.92,
At failure: 108.0 = B [0 + A x 475.0] = 0.92 x A \303\227
475.0, A = 0.247.

4.3.7 ln-situ shear strength: shear vane test

In soft and saturated clays where an undisturbed specimen is difficult to obtain, the
undrained shear strength is measured using a shear vanc tcst. A diagrammatic view of the
apparatus is shown in Figure 4.20. It consists of four thin mutually perpendicular blades

Figure 4.20. Shear vane apparatus.)))

132 Soil Mechanics: Basic Concepts and Engineering Applications

connected to a central rod where the height H is twice the overall diameter D. Conunonty
used diametersare 38, 50 and 75 mm. The vane is pushed into the soil either at the ground

surface or at the bottom of a borehole until totally embedded in the soil (at least 0.5 m).A
\302\260
torque T is applied and the vane is rotated at the slow rate of 9 per minute. As a result,
shear stressesare mobilized on all surfaces of a cylindrical volume of the soil generated by
the rotation. The maximum torque is measured by a suitable instrument and equals the
moment of the mobilized shear stress about the central axis of the apparatus, From this,

and after suitable rearrangement, we obtain:

T
cu
=
(4.28)
\177DZ(H /2 + D/6)
If the test is carried out with a rapid rotation, then the result may indicate the remoulded
undrained shear
strength and could be used as a measure of sensitivity. The sensitivity in
terms of comparison between undisturbed and remoulded samples is defined as the ratio
of cu obtained from the standard rate to the Cu obtained from the rapid test.

4.4 STRESS-STRAINBEHAVIOUR OF SANDS AND CLAYS

4.4. l Stress-strainbehaviour of dense and loose sands

Figure 4.21 shows typical results of direct shear tests in dense and loose sands and, based
on these results, the following observations can be made. In dense sand the shear strength
increases with the shear displacement (relative displacementbetween the and the
lower
upper box) to a maximum value of Zmax
=
\"tf,
and then decreases to a critical value of Xcr
that is approximately a constant for each type of sand. The term critical state is used to
describe the state of the soil at this point. At the critical state the sample is sheared while
having a constant volume. If further shear displacement is applied, the shear strength de-
creases to a residual value \"\177and shear planes will be fully developed within the sample.
Initially, of the specimen
the volume decreases slightly, and then increases with further
shear displacement. This dilation can be explained by plotting the vertical movement (z)
of the loading plate, or the normal force N of Figure 4.8(a), against the shear displacement
(x). The rate of volume change is represented by the dilation angle \177, which is the slope
(\177z/Ax) of the resulting curve. The upper half of the shear box diverges from its horizontal

Shear Displacement
(c))))
 Shear Strength of Soils and Failure Criteria 133

directionto follow the dilationAs the increasing shear displacement

angle. moves the
sample towards its residual
state, the dilation angle decreasesto zero. Dilation is attribute
to interlocking of the sand particles that have to be subsequently loosened to allow the

shear plane to develop. This (volume change) behaviour may be illustrated by plotting
void ratio (or specificvolume) against shear displacement void (Figure 4.2 l(c)). The
ratio
of the sample increasesfrom its initial low value to a higher critical value, which is ap-
proximately a constant for each type of sand. Note that in the calculation of the void ratio
(while shearing) the thickness of the
specimen is assumed to be the thickness of the failure
zone. As the failure zone is more concentratedimmediately above and below the horizon-
tal failure plane, standard computations may underestimate the critical void ratio for dense
sands. The magnitude critical void ratio, or the equivalent
of the internal friction angle,
depends on the level of normal stress applied to the sample. When the normal stress is in-
creased to very high levels, the internal friction angle corresponding to the peak strength
decreases and approaches the critical value, thereby creating a two-stageMohr-Coulomb
failure criterion. In loose sand, the shear strength increases with the shear
displacement
and, without passing through a peak point, approaches its critical value at which it remains
approximately constant. The volume of the specimen gradually decreases to a certain
value and remains
approximately constant thereafter.
Similar patterns of behaviour for dense and loose sands are observed in triaxiai com-
\302\260
pression tests. However, the angle of internal friction related to the peak point is 4 \302\260-
6

less than that obtained in a modified direct shear test (Arthur & Aysen, 1977) in which the
modification facilitates the generation of shear stressesat the end planes. This difference
is mostly as a result of axisymmetric conditionsin the triaxial test and the angle between
direction of the deposition of the particles with the principaI stress directions.

Example 4.12
The results shear test on a sample of dense medium
of a direct size dry sand are tabulated

below. The
sample reached equilibrium under a normal stress of 100 kPa and its initial
void ratio under this load was 0.51. If the height of the sample was 20 mm, (a) plot the
variation in shear stress and vertical displacement against the shear displacement, (b) plot
the variation in void ratio against the shear displacement for:
h = thickness of the sample = 20 ram, h = thickness of the failure zone = i0 ram.

Reading Shear Vertical Shear Reading Shear Vertical Shear

points disp.(rmn) disp.(mm) stress (kPa) points disp.(mm) disp.(mm) stress(kPa)

1 0.25 0,00 25 11 2.75 1.01 60

2 0.50 0,08 50 t2 3.00 I. I 1 56

3 0,75 O, 19 68 13 3,25 1.20 54
4 1.00 0.29 78 14 3.50 1.28 51
5 1\303\26725 0.39 83 15 3.75 1.34 50
6 1.50 0.50 84 16 4.00 1.39 49
7 1.75 0.60 80 17 4.25 1.42 48
8 2.00 0.71 74 18 4.50 1.45 48
9 2.25 0.81 69 19 4.75 1.47 48
I0 2.50 0.91 64 20 5.00 1.48 48

Solution:
(a) Figures 4.22(a) and 4.22(b) show the variation in shear stress and vertical displacement
against shear displacement from which:)))
134 Soil Mechanics: Basic Conceptsand Engineering Applications

100

80

60

40
20
0
1 2 3 4 5 1 2 3 4
Sheardisplacement (mm) Shear diplacement (ram)

(a) (b)

0.75
0.7

0.65

0,6

0.55
0.5 \177=
0 1 2 3 4 5
Sheardisplacement (ram)

Figure 4.22. Example 4, [2,

tanO (or 0\") = (maximum shear strength

= 84
kPa) / (normal stress = 100 kPa) = 0.84,

O = 40 From
\302\260. Figure 4.22(b) and for reading points 4 and 12:

tan\177
= (1,i! mm- 0,29 /
mm) (3.0 mm - 1.0 =
mm) 0.41, \177 = 22.3 \302\260.
(b) Sample calculation for shear displacementof i mm:
Void ratio at any stage: e = e0 + Ae = e0 + zA / hsA = eo + z / hs
where e0 is the initiaI void ratio, z is the vertical displacement, A is the cross-sectional area
of the shear box, and hs is the equivalent height (or thickness) of the solids.

Reading Void ratio Void ratio Reading Void ratio Void ratio
points h=20mm h=[0mm points h=20mm h=10mm

0.510 0.5 I0 I I 0.586 0.662

l 0,5 l0 0,5I0 12 0.594 0.678
2 0.516 0.522 13 0.600 0.69I
3 0.524 0.539 I4 0.607 0.703
4 0,532 0,554 15 0.61I 0.712
5 0.539 0.569 16 &615 0.720
6 0,548 0,585 17 0.617 0.724
7 0.555 0+600 18 0,619 0.729
8 0.564 0+617 19 O621 0.732
9 0.57 [ 0+632 20 0.622 0.733
[0 0.579 0.647)))
 Shear Strength of Soits and Failure Criteria 135

Calculation ofh\177: e0
= 0.51 = Vv / Vs. or Vv = 0.51V]\177. Substitute Vv in: V= +
V\177. Vv:

hs -\177 hs - V / ( 1.5 tA) = h \303\227

Vs = V / 1.51 = A(area) \303\227 A / 1.51 A = h / 1.51,

where h is the height or thickness of the sample. For h = 20 mm we have:

e= eo+z/hs =0.51 + (1.51/20) z. z=0.51+0.0755
If we assume the volume change is concentrated within the 10-mm zone centred on the

failure plane, then: e(h = lO mm)

= 0.51 + 0.151 Z.

For a shear displacement of 1.0mm, z = 0.29 turn:

e = 0.51 + 0.0755
\303\227
0.29 = 0.532,
e(t,_ 10 m.m) =0.51 + 0.15t 0.29
\303\227 = 0.554.

Void ratios for h = 20 turn and h = 10 mm are tabulated and are shown in Figure 4.22(c).

4.4.2 Stress-strain behaviour of ctay

Figures 4.23 and 4.24 present the results of drained and consolidated-undrained triaxial

tests on fully saturated clay soils. In drained tests, a normally consolidated clay specimen
contracts under
increasing deviator stress until a constant volume is reached. The constant
volume underwhich shear planes start to generate is called the critical volume. At this
stage, the angle of internal friction is slightly greater than the residual internal friction an-
gle but equal to or less than that of the peak point (if it exists). In general the failure enve-

Iope passes through the origin. An overconsolidated clay, however, expands as it is

sheared and, after passing a peak point, approaches a constant volume. For this soil, the
failure envelope indicates the presence of some cohesion, the magnitude of which depends
on the type of soil. The magnitude of the expansion depends on the overconsolidation ra-
tio (the maximum effectivestressonceapplied to the soil in the past, or preconsolidation

pressure divided by the present effective stress) and increases as the overconsolidation ra-
tio increases. In determining the magnitude of the confining pressure applied in a triaxial

test on samples of both normally consolidated and overconsolidated clays, consideration

must be given to the field conditions. Care must be taken in constructing the failure enve-
lope to distinguish test points that are situated to the left or right of the preconsolidation

pressure. When the test points are scattered on both sides of the preconsolidation pressure,
the best fit for the failure envelope may not be the best answer.
In the consolidated-undrained test, water is not permitted to flow in to or out of the
sample, and positive or negative pore pressures are generated as the sample is sheared.

- G
\302\260'3 \1771 3

\177__

% \177\177\177-\177-,
\177\"
Compression \177v Compression

(b)

Figure 4.23. Drained triaxial tests on cohesive soils: (a) normally consolidated clay, (b) overconsoli-
dated clay.)))
136 Soil Mechanics: Basic Concepts and Engineering Applications

o\"
3

(b)

Figure 4.24. Consolidated-undrained triaxial tests on: (a) normally consolidated clay, (b) overcon-
solidated clay.

In a normally consolidated clay an increase in deviator stress creates a tendency for the
soil to contract and push water out of the sample, but flow is not permitted, positive
as

pore pressureis generated in the pore water. In an overconsolidated clay, however, after
an initial increase in pore pressure, the soil tends to expand and therefore draws water in,
but, as water is prevented from flowing, a decrease in pore pressure will be produced that
may develop into a negativepore pressure.The magnitude of this negative pore pressure
dependson the overconsolidation ratio and increases (in negative values) as the overcon-
soIidation ratio increases. Similarities between the stress-strain behaviour of loose sand
and normally consolidated clay on the one hand, and between dense sand and oveconsoli-
dated clay on the othersuggest that there may exist a unique
hand, relationship relating the
state of the stress volume change properties,and a general failure criterion for soils.
to the
In an unconsolidated-undrained triaxiaI test carried out on a fully saturated clay, water
flow is prevented in to or out of the sample at every stage of the test. Any applied confin-
ing pressure or3 is then resisted solely by the pore water and the incremental effective
stress remains zero. The increase in deviator stress is then resisted by the undrained cohe-
sion cu that is an inherent strength characteristic of the soil. Frictional resistance cannot be
mobilized, as the effective stress remains zero. Consequently, the failure envelopes for
both total and effective stresses are the same, as is shown in Figure 4.25. This means that
the strength is the same regardlessof the magnitude of the applied confining pressure.

Example 4. t3

Using the data of Example 4.9 plot the variation of deviator stress and pore pressure
against axial strain. Show the stress - strain behaviour in the normalized coordinate system

Figure 4.25. Unconsolidated-undrained triaxial test on a saturated clay.)))

 Shear Strength of Soils and Failure Criteria 137

1000 1.6
t\177 = 45CIkPa 1.4
800 \177 3

600
\"\177
0.8
= 15C kPa -\177_
400 \177
'\177
0.6
\177
0.4
200 leq-\302\260\"
lu 0.2
OI
0 0.1 0.2 0.3 0.4
-200
\177ial strain
0 0.1 0.2 0.3 0.4 \177

Axial strain (b)

,51

(a)

Figure 4.26. Example 4..13.

of El, q\" /p\" and discuss the results.

Solution:
The variations of deviator stress
and pore pressure against axial strain is shown in Figure
4.26(a). Figure 4.26(b) shows the normalized stress-strain behaviour in El, q'/p\" coordi-
nate system. The parameter q' is calculated by multiplying t' by 2; the parameter p' is cal-
culatedfrom Equation 4.16. The generation of negative pore pressure in the first sample -
= 150 kPa, \302\260
with {\1773 a peak strength O\" of 33.9 a residual
\302\260, 0p' of 28.9 (calculations not
shown), zero effective cohesion,and the pattern of stress paths shown in Figure 4.17(b) -
shows that the soil is a lightly to medium overconsolidated fine silt.

4.5 CRITICALSTATE THEORY

4.5. t Isotropic compression

From the discussion 4.4 it is evident that, as a soil is sheared,the elastic-plastic
of Section
deformations the state of the soil into a condition of perfect plasticity
transfer in which

plastic shearing continues with constant effective stress and volume. In soil mechanics this
condition is known as critical state. Critical state theory proposes a relationship between
the state of stress and the specific volume v while shearing from an initial state to the
critical state.
The consolidation stage in consolidated-undrained and drained triaxial tests is repre-
sented by point A in the p', v coordinate system, as shown in Figure 4.27(a). At this point,
P'A
= (0,1 20'3) / 3 0\"3 t\1773. Any further increase in the confining
+ = = pressure will induce
a reduction in volume, which is represented by the point B. The state of the sample sub-

jected to an isotropic compression or confining pressure {\1773is defined by a relationship be-

tween v and p' called the normal compression line (NCL), as shown in Figure 4.27(a).
This normal compression line is established by means of a triaxial compression test. If at

any point on the NCL, say point B, the sample is unloaded (arc BC), the volume recovers)))
 138 Soil Mechanics: Basic Conceptsand Engineering Applications

V V

'\177CL

P\177A P'B

(a) (b)

Figure 4.27. Isotropic compression in v, p\" and v, In p' coordinate systems.

slightly and the sample expands. Reloading causescompressionand the state of the soil
moves towards the normal compression line on arc CD. The state of the soil on BC or CD
is termed as overconsolidated. The deformation behaviour of the soil along these lines can
be assumedto be elastic whereas, on the normal compression line, the soil deformation

bebaviour is plastic. It is seen that the NCL is a boundary to this state, and that any condi-
tion to the left of the NCL when moving from A to D cannot exist.

Figure 4.27(b) represents the normal compression line in v, In p' coordinate system.
Experimental results show that the normal compression line can be approximated without

serious error by a line defined by:

= N-XIn
v p' (4.29)
where the slope of the
X is normal compression line and N is the magnitude of v at

p'
= lkPa. The expansion Iine BC and recompression line CD can be idealizedby just one
line with a slope of K, which implies that during unloading and reloading, expansion and
recompression are equaI and points B and D coincide. The equation of this line may be

presented in a similar way as Equation 4.29:

v = -
v\177 KInp' (4.30)
where K = -(v - v - In and vK is the of v at p'c = i kPa.
c, t))/(In PC PL)) magnitude

Note that vK is not a constant for the soiI and its magnitude depends on the magnitude of

P't). Alternatively:

(V-Vc)/(lnp'-lnpc -K, or: (V-VD)/(Inp'-lnpo)=-K (4.31)

For clay soils the normal compression line is establishedat relatively low to medium lev-
els of stress. For stress level should
sand, the be more than 700 k_Pa in order to approach
the relevant NCL (Atkinson & Bransby, 1978).Recompression at a moderate stress is rep-
resentedby a horizontal line that indicates no volume change during this process. At the
same time, there is no relationship between the state of stress and the initial volume. For

example, a very loose sample of sand that is prepared by pouring the sand into a container,
could turn into a dense sand if it is submerged in water and the water is allowed to drain.
Whilst the process induces no serious stresses in the sand particles, there is a marked
difference in the void ratios. In general, the behaviour of sand at higher stresses is similar
to the behaviour of clay subjected to lower isotropic compression.)))
 Shear Strength of Soils and Failure Criteria 139

One-dimensional consolidation is represented by the line VL in Figure 4.27(b). The

slope of this line is similar to the of the normal compression line and it is convenient
slope
to assume that they are equal. As a result, the equation of the one-dimensional consolida-
tion line is:

v = fi'-)vln p' (4.32)

where )v is the slope of the normal compression line and g is the magnitude ofv in an oe-
dometer test with p'= ikPa. Consolidation test results are often
presented log \177'in a e,

coordinate system where e is the void ratio and \177 is the effective vertical stress appliedto
the sample. The compression index Cc is the slope of the linear portion of e-log \177 plot

and it can be shown that Cc = In (10) x \177.= 2.3025 \177,.The slopes of the expansion and re-

compression lines in this case have the same ratio (2.3025) to the expansion and recom-

pression lines in normal compression. These lines represent elastic deformation which is
recoverable while the deformation on e-log o\" line is plastic and unrecoverable.

An element of soil in the field may undergo eIastic and plastic deformation depending
on the level of the loading, the coordinate of the element, its initial state of stress and spe-
cific volume. Critical state theory combines the parameters, L, \177:,N, and n with the state of

stress and gives an elastic-plastic model for the soil behaviour while moving from a given

initial state to the critical state.

Example 4.14

Results obtained from an isotropic compression test on a specimen of saturated clay are
tabulated below. The volume of the sample and the moisture content at the end of the test
are 76.80 ml and 30.6% respectivelywhiIst the specific gravity of solids is 2.7. Calculate
\177.,N, and K and plot the normal compression line. What are the overconsolidation ratios at
test points i and 7'?

Test Confining Volume Test Confining Volume

points pressure (kPa) change (rnl) points pressure (kPa) change (ml)
1 15 0.00 5 400 14,25
2 50 1.75 6 50 11.20
3 100 5,50 7 15 9.40
4 200 9.90

Solution:
Calculatethe void ratio and specific volume correspondingto each loading:
For a saturated soil the void ratio is (Equation 1.15): e = w Gs = 0.306 x 2.7 = 0.826,thus
v7 = 1 + e7 = 1.826. Calculate the volume of solids:
Vv + Vs = 76.8 =
ml, e7 Vv Vs / = 0.826 or Vs
= 76.8 / (i + 0.826) ---42.06 ml.

The void ratio and specific volume corresponding to each load is calculated as follows:
e6
= e7 + AV / Vs.
= 0.826 + (9.4 - I 1.20)/ 42.06= 0.783, v6
= 1.783,

e5 = e7 + k.V / V.\177.
= 0.826 + (9.4 -- i4.25) / 42.06= 0.71I, v5 = 1.711, and similarly:

v4 = 1.814,v3 = 1.919, v2 = 2.008, vl

= 2.050.

The results are shown in Figure 4.28. Points i and 2 make the recompression line, while
points 3, 4, and 5 represent the normal compression Iine. The line of best fit

gives: \177,- 0.15, \177= 0.035. Substituting the \177.and v values of points 3, 4, and 5 into Equa-
tion 4.29 we obtain an average value of 2.61 for N.)))
140 Soil Mechanics: Basic Concepts and Engineer#tg Applications

2.1

1.9

1.8

1.7

1.6
3 4 5 6
In p'

Figure 4.28. Example 4.14.

In order the overconsolidation ratio at point 1, we need to know

to calculate the maximum

isotropic effective stress (P'c) that has been applied to the soil in the past. The preconsoli-
dation pressure is the effective stress at the intersection point of line 1-2 with the NCL.
Equation of line 1-2: v
- 2.05 =-
0.035(lnp' - In 15.0)
Equation of NCL: v = 2.61 - 0.15 In p'. Solving for p':
p' =57.1 kPa, 0CR=57.1/ 15.0=3.8.
At point 7: OCR = 400.0 / 15.0= 26.7.

4.5.2 The critical state line

The critical state line (CSL)is a spatial curve in the v, p', q\" coordinate system (Figure
4.29) and is located on a surface called the state boundary surface (SBS). This surface in-
tersects the v-p' plane at the normal compression line. It bounds all possibte states during
loading and relates the state of stress to the specific volume at any stage of loading while
moving the state of the soil from a specified initial conditions towards the critical state
line. Therefore,if the equation of the SBS is known then evaluation of the elastic and plas-
tic strains is possible at any stage of the loading. In this book we will concentrate on the
basics of the critical state theory in terms of the definition of the CSL and its application.
Calculation of strains, and distinguishing elastic strains from plastic strains, that require
the derivation of SBS and other mathematical operations are beyond the scope of this
book. Based on the critical state theory a comprehensive model for lightly consolidated

clay was developed at Cambridge University during the 1960s, and subsequently became
known as original Cam clay model (Roscoe & Schofield 1963).This model proposed a
mathematical equation for state boundary surface that included a logarithmic spiral-shaped
yieId surfacein the p', q\" coordinate system. This yieldsurface was later modified (Roscoe
& Burland, 1968) to an elliptical surface, and this became known as modified Cam
model
clay model. Carter et al. (1982) extendedthe modified Cam clay model to predict the be-
haviour of a soil under the cyclic loading. The Cam clay based models have been applied
successfully to geotechnical engineering problemsusing numerical methods. Applications
of the finite element method to the critical state theory can be found in many texts and re-
searchpapers(e.g.Britto & Gunn, 1987).)))
 Shear Strength of Soifs and Failure Criteria 141

q\302\242

p,

/CSL

Figure 4.29. Critical state line in v, p', q\" coordinate system.

Projection of the CSL onto the p'-q' plane (Figure 4.29) defines the state of stress at the
critical state and is a line that passes through the origin. The gradient of this line is defined

as M, which is one of the critical state parametersof the soil, and its equation is:

q'= Mp\" (4.33)

This means the strength of the soil depends solely on the normal effective stress and that it
has no cohesion. Thus the corresponding critical friction angle can be found by equating
M to the first term of Equation 4.17. The volumetric behaviour at the critical state is de-
fined by the projection of the CSL onto the v-p' plane (Figures 4.29 and 4.30(a)). The ini-
tiaI position is located on the v-p' plane shown by the points C, L, and H. Point C is on the

NCL and represents a normally consolidated soil. Point B is located between the NCL and
the projection of the CSL, and represents lightly consolidated soil. Point H is located be-
tween the v axis and the projection of the CSL, and represents a heavily consolidated soil.

v v

vH _H_ ..... _L\"

', ', i
CSL\177 p, In p')))
P'H P'L P'B P'H P'L P'C
(a) (b)
Figure 4.30. Projection of the CSL onto the v-p' and v-In p' planes,
142 Soil Mechanics: Basic Concepts and Engineering Applications

Depending on the stress path during loading, the initial point will move towards the criti-
cal state line while being inside or on the state boundary surface. The projection of the
CSL onto the v, In p' coordinate system (Figure 4.30(b)) can be approximated by a line

parallel to the NCL and located on the permitted side of it:

v = F-Xtn p\" (4.34)

where F is the critical specific volume at p' = i kPa and X is the slope of the line equal to
the slope of the NCL.

4.5.3 Triaxial tests on normally consolidated saturated clay

The stress paths

and undrained tests on normally
for drained consolidated clay are shown
in t\177igure 4.31. Point C
represents the initiat state under a confining pressure c\1773.The stress
path CD of the drained test is on a plane that passes through point C. It is perpendicular to
the p'-q\" plane and has a gradient of 3 vertica\177 to t horizontal with the v-p' plane. Point D
representsthe critical state and is the intersection point of the plane describedabove with
the CSL. The stress path is on the state boundary surface and is defined by its projections
on p'-q\" and v-p' (or v-ln p') p\177anes. Projections of the stress path onto the p'-q\" and v-p'
(or v-In p') planes are p'C-Dp'q\" and
C-Dp'v respectively. The stress path CU of an
undrained test is on a plane that passes through point C and is perpendicular to the v-p'
\302\260
plane,making an angle of 90 with the v axis to ensure a constant votume during the test.
Point U represents the criticaI state and is the intersection point of this plane with the CSL.

qP

q'CD ..........

q'CU

I:l'c
P'\177u

v (a,')

v
C

VCD
' -\177
] i i\177p '\177
; CSL.,_ I I '
\177 p, \177 Inp'
P'CU P'c P'CD P'CU P'C P'CD
(b) (c)
Figure 4.31_ Triaxial tests on a normally consolidated saturated clay.)))
 Shear Strength of Soils and Failure Criteria 1,43

The path CU is located on the surface of the state boundary, meaning that similar paths
initiating from the different points on the
compression line construct the surface
normal of
the state boundary. The projections of the stress path on p'-q\" and v-p' (or v-In p') planes
are p'c-Up'q' and C-Up'v respectively. As the effective stress path is on a constant v plane
therefore its projection on the p'-q' plane is identical to its position in the space. In the
case of the original Cam clay model the following state boundary surface is used:
'
F-v -L 1
n.p )
q'=Mp'(l+ (4.35)

This surface meets the v-p\" plane on the normal compression line v = N- L in p\" and is
tangent to the q'-v plane along the v axis where q\" =
p'
= 0, Equation 4.35 implies that:

N -I
\177
= L - \177 (4.36)

The equation of the undrained stress path can be obtained by intersecting the SBS with a
constant specific volume plane:

,\177-K p

where P'C is the initial mean effective stress on the norma] compression line (isotropic
compression or isotropicconsolidation pressure) corresponding to the constant v

For most normaliy consolidated clays Equation 4.37 represents the projection of the state
boundary surface on the p'-q' plane. This surface (Equation 4.37) is called the Roscoe sur-
face (after Professor K HRoscoe, Cambridge University, UK). The relationship between
the mean effective stress at the critical state P'CUand the initial mean effective stress P'C
can be established by combining Equations 4.37 (with replacing p'cu for p') and 4.36:

N -F
=
in( p._\177p,c_._c (4.38)
)
Pcu \177,

Undrained shear strength cu. The undrained shear strength at the critical state can be de-
termined from the following equation:
cu =
(or I
- cr
3 )/2
= (cr\177 - \177g)/2
=
q'/2 --
(Mp'cu) /2 (4.39)
where q' is the deviator stress at the critical state. Substituting p'cu from Equation 4.38 the

undrained shear strength can be related to the isotropic compression or isotropic consoli-
dation pressure P'C according to:

c, = (Mp\" C/2)exp[(F- N) / Z ] (4.40)

As the specific volume is a function of the isotropic compression p'c, then Equation 4.40

can be written in terms of specific volume by repJacing p'c from Equation 4.29. Hence:

c,, = (M / 2) exp[(F
- v) / \177,i (4.41 )

In a fully saturated clay the void ratio e = wGs (w is the moisture content and Gs is the

of solids). Therefore v = 1 + e = i + wGs, and for v in

specific gravity substituting Equa-
tion 4.41 we obtain:

cu = (M / 2) exp[(F' - 1 -
wG.,. ) / \177,] (4.42))))
 144 Soil Mechanics: Basic Conceptsattd Engineering Applications

The undrained shear strength may be expressed in terms of the one dimensional consolida-
tion parameters P'IC (one dimensional consolidation pressure)and fi by combining Equa-
tions 4.32 and 4.41:

cu =
(Mp\177 c / 2) exp[(l\"
- \177')/ L] (4.43)

Example 4.15
Two triaxial compression tests identical specimens of a particular
were carried out on two
clay whose diameter was 38
height mm The and
first sample 76 mm.
was isotropically
consolidated under 300 kPa, and then subjected to drained test. The critical state was
achieved under a deviator stress of 360 kPa and the volume change was 4.4 ml. After

completion of the test, the sample was oven dried and the mass was measured as 145.8 g,
The second specimen was isotropically consolidated under 300 kPa, and then subjected to
an undrained test. The critical state was achieved under a deviator stress of 152 kPa. De-
termine the values of M, L, F, N and the ultimate pore pressure generated in the second
specimen. Gs = 2.72.
Solution:

p'
= (o\"i + 2c(3) / 3 = (\177'1
-
0\"3 + 3c\177'3) / 3 = \177'3 + q' / 3,
p'(drained test) = 300.0+ 360.0 / 3 = 420.0 M = q\"/p\" = 360.0 / 420.0 = 0.857.
kP\177
Initial volume of the specimen: x (3.8) x
Vo
= 7.6 / 4 = 86.193 ml,
r\177

Volume at the end of the test = 86.193- 4.4= 81.793

ml,

Volume = 145.8 / (2.72 x 1.0)= 53.603

of solids ml,

Volume of voids = 81.793 - 53.603 = 28.19ml,

Void ratio at critical state = 28.19/ 53.603= 0.526,

Specific volume at critical state = vf = i + 0.526 = 1.526. The point with v = 1.526 and

p'
= 420 kPa is on the projection of the CSL onto the p'-v plane.
Initial void ratio = 0.526 + 4.4 / 53.603 = 0.608.
v0
= I + 0.608 = 1.608 (at the start of both drained and undrained tests but is constant dur-

ing the undrained test).

At the end of the undrained test: p\" = q' / M = 152.0/ 0.857= 177.4 kPa.

The point with v = 1.608 and p'

= 177.4 kPa is on the
projection of the CSL onto the v-p'

plane. Having two points on this line the equation of the CSL in the v, Inp' coordinate sys-

tem can be found:

=
9\177 (1.608
- 1.526)/(in 420.0 - In 177.4)
= 0.095.

Substituting coordinates
the of one of these points into Equation 4.34:
1.608 = F - 0.095 In 177.4, F = 2.1. Therefore the equation of the CSL in the v-p' plane is:
v = 2.i - 0.095In p'. The initial states of both samples (p\"
= 300 kPa, v --- 1.608) are on the
normal compression line. Using Equation 4.29:1.608 =N-0.095 In 300.0, N= 2.15.
Hencethe equation of the NCL is: v = 2.15 - 0.095 In p'.
Using Equation 4.16:\177'3 177.4 152.0
= - / 3 = 126.7 kPa. u = 300.0
- 126.7 = 173.3kPa.

4.5.4 Triaxiaf tests on lightly overconsolidated saturated clays

Projections of the stress paths on the p'-q', and v-p' (or v-In p') planes for two specimens
of lightly overconsolidated clay (point L) are shown in Figure 4.32. The stress paths in the

space are located insidethe state boundary surface and in the undrained test the stress path)))
 Shear Strength of Soils and Failure Criteria 145

q'LD

q'LU

v
L

VLD

I I I
\177 In p'
P'LU P'LU P'L P'LD

(b) (c)
Figure 4.32. Triaxial tests on a lightly overconsotidated saturated clay.

is identical to its projection onto the p'-q' plane. If the initial position of point L is on the
projection of the CSL onto the v-p' plane, then the undrained stress path on the p'-q' plane
is a vertical line having a constant effective mean stress equal to P'LU. The preconsolida-
tion pressure for this initiaI state is on the normal compression line and itbe showncan
that in the original Cam clay model the ratio of the effective mean stress at this point to
the initial effective mean stress at L is equal to e (the base of natural logarithms)
= 2.72.
This means that the overconsolidation ratio OCR at this point is 2.72. Note that at any
point on the normal compression line OCR = 1.

Example 4.16

A specimen or\" the soil of Example 4.15 is consolidatedunder 500 kPa and then unloaded
to 200 kPa. The volumetric strain due to expansion is 2,64%. Determinethe range of the
consolidation pressures to prepare samples of void ratio of 0.62 and the formulation of the
unloading.
Solution:
(a) Calculationof 4.29:Vs00 = 2.15-0.95 In 500.0
K: from = 1.560.
Equation
- 1.560) / 1.560 = -0.0264 -\177 v
2o0 v 50o) / v 500 = -(v 2o0
\177 = -(v - = 1.601.
v 200
K = -(1.60[
- 1.560)/(In 200.0 - In = 0.045. Check if N - F = X - \177::
500.0)

2.15 -2.1 = 0.095 -0.045 = 0.05, thus the Cam clay model can be applied.
The initial states of the specimens are within P'A to p'/3, as shown in Figure 4.33.
Applying Equation 4.34:1.62 = 2.1 - 0.095In -\177 = 156.4 kPa.)))
p\177 p\177
146 Soil Mechanics.\" Basic Concepts and Engineering Applications

v =1.62 .......
A
......L B
....

, I I 1 \177 i
\177_ In p'
P',\177 P'c P'8 P'v P'c
Figure 4.33. Example 4.16.

From Equation 4.29:1.62 = 2.15

- 0.095In PB
-> PB = 264.8 kPa.
To find the mean effective stress at point C (Figure 4.33) we construct two equations
based on the definition of K and
-K = -0.045 = (1.62-VC)/(ln 156.4- In p\177),

-\177, = -0.095 = (1.62 - v

C )/(In 264.8 - In
P\1777), solving for vC and P'C.

p'c 425.3 kPa (note that the Iog), vc

425.3 = / 156.4 = 2.719 = baseof natural 1.575.
To amount
determine of load after
the unloading from NCL, we draw line LV parallel to
AC, where L is the state of a lightly overconsolidated specimen.
-K = -0.045= (v L -v v )/(In p\177 - In p\177 ), replacing vv from NCL:
-0.045 = [1.62 - 0.095
- (2.15 In p{/)]/(In p\177,
- In
p\177/),
after rearranging and simplifying:
, \342\200\242 ,
Inp L =ll.77-1.111npV
-\177
p \177.,=130323/ pv ,
where: 156.4kPa <-P'L
< 264.8 kPa and 264.8 kPa \177P'V-< 42.5.3 kPa.

4.5.5 Triaxial tests on l\177eavily overconsotidated saturated clays

Projections of the stress paths onto the p'-q', and v-p\" (or v-lnp') planes for two specimens
of heavily overconsolidated clay (point/4) are shown in Figure 4.34, The stress path of the
drained test on the p'-q' plane has a slope of 3 vertical to 1 horizontal, and passes beyond
the critical state line (point D) approaching point L which represents the peak deviator
stress. Ideally. if straining is continued, the state of the soil must descend to point D.

However, any additional strain beyond L may produce failure planes, thereby preventing
attainment of the critical state. The line SU is a part of the state boundary surface corre-
sponding to the initial specific volume and is called the Hvorslev surface (after Professor
M J Hvorslev). This surface is the envelope for the peak strength of drained triaxial results
having the same specific volume at peak irrespective stress conditions.With
of the this

definition a series of tests with different void ratios at failure (peak) will yield parallel en-
velopes, The ideal path for the undrained test is to approach the intersection of the Roscoe
surface and the CSL on the Iqvorslev surface. However, the development of failure planes

may terminate the path at the left of point U and he.ice the critical state may not be
achieved. The equation of the Hvorslev surface in the v, p', q' coordinate system is:)))
 Shear Strength of Soils and Failure Criteria 147

VHD
.... .....................
\177r
V_H_O_

V
H

P'H P'CU P'C P'H P'CU P'C

(b)

Figure 4.34. Triaxial tests on a heavily overconsolidated saturated clay.

q'= Hp'+(M
- H)exp[X(F- v)] (4.44)

where H is the Hvorslev surface on the p'-q\" plane. The line OS (with a slope
slope of the
of 3 vertical to represents a drained test with zero confining pressure. This
1 horizontal)

line eliminates the tensile stress on the Hvorslev surface (no tension cut oft) and thus the
complete section of the state boundary surface on the p'-q' plane is OSUC.
In order to evaluate the overconsolidation ratio for a constant specific volume, consider
Figure 4.35 in which L represents an overconsolidated clay and OL is the expansion line.
The two points C and O are on the normal compression line, thus:

v C -v o =
),(In p\177)
-In
PC) (4.45)

On the expansion line the change in specific volume is defined by:

,CSL NCL

' \177 \177-- Inp\"

P'L P'C P'O
Figure 4.35. Evaluation of overconsolidation ratio.)))
!,48 Soil Mechanics: Basic Concepts and Engineering Applications

v L-v o =v C -v o = K(lnPO-lnpL) (4.46)

Equating equations 4.45 and 4.46 and considering the definition of the overconsolidation

ratio OCR, we find:

/ r c
-v
OCR =
\177
= (-(\177--)
\177\"
= exp(V o.) (4.47)
K
PL PC

4.5.6 Three-dimensional and normalized state boundary surface

Figure 4.36(a) shows the three-dimensional state boundary surface in the v, p', q' coordi-
nate system. The section OSUC of Figure 4.34(a) can be normalized in terms of the pre-
consolidation pressureto represent the unique section of the state boundary surface in the
-
P\" / P'c q\" / p'c plane as shown in Figure 4.36(b). During the progress of the drained test,
the stress path intersects different sections having a different P'c.
with constant v, each
The state each intersection must
of the stress at against the corresponding be normalized
preconsolidationpressure,and it can be shown that the resulting path will be the same as
the normalized Rosoce surface for the undrained condition. To describe samples that lie
between the NCL and the CSL on the v-p\" plane, the term wet of critical is used.The
maximum difference in the void ratio within this area is N- F and, clearly, samples in this
area have a greater moisture content than samples below the CSL that are dry of critical.

The stateswithin each group have similar stress strain patterns.

Example 4.17

The data obtained from triaxial tests on two specimens of normally consolidated clay are
shown in table below (q', u for undrained, q', ev for drained) where the last row represents
the critical state.

CSL Roscoesurface
Hvorslef\177

q'/P'c

CSL
Hvorslef U \177

'/
\177,
\177 D\177of. iWet\302\260f
\177 criiical critical
\177

/ / Notension

v P'/P'c

(\177)

Figure 4.36. (a) Three-dimensional state boundary surface, (b) normalized state boundary surface.)))
 Shear Strength of Soils and Failure Criteria 149

Test Undrained: 0\"3

= 200 kPa Drained: c\1773= 150 kPa
points q' (kPa) u (kPa) p' (kPa) q' (kPa) \302\242v,\177 p\" (kPa)

I 0.0 0.0 200.0 0.0 0.00 150.0

2 27.8 29.3 179.9 15.6 1.10 155.2
3 52.4 57.4 160.0 33.9 2.31 161.3
4 73.2 84.4 140.0 55.1 3.57 168.4
5 82.1 97.3 130.1 80.3 4.99 176.8
6 89.9 110.0 119.9 112.2 6.63 187.4
7 96.4 122.1 lI0.0 149.5 8.42 199.8
8 101.7 133.9 100.0 200.0 10.53 216.7
9 103.9 140.2 94.4 260.4 12.74 236.8

The moisture contents at the end of the tests are 31% and 24.21% for the undrained and
drained tests respectively. (a) Plot the stress paths for both undrained and drained tests in
the p'-q\" plane, (b) evaluate the normal compression and critical state parameters, (c) plot
the normalized stress paths for both samples on p\" / P'c - q\" / p'c plane. Gs = 2.71.

Solution:

(a) From Equation 4.16: p' (undrained test)

= 200.0 - u + q' / 3.
For the drained test: p' = 150.0 + q' / 3. The values ofp' for both tests are tabulated above

and both stress paths are shown in Figure 4.37(a).

(b) In the test and at the critical state M = q\"/p\" = 103.9 / 94.4 = 1.1.
undrained
CheckM from the drained test results: M = q\"/p\" = 260.4 / 150,0 = 1.1.
Calculate the final void ratios for both tests:

e9 (undrained) = w G\177= 0.31 x 2.71 = 0.840, e9 (drained) = 0.2421 x 2.71 - 0.656; thus:

v9 (undrained) = =
1.840, v9 (drained) 1.656.
The initial void ratio (drained) can be calcuIated using the definition of voIumetric strain:

\177v =AV/Vo =-(v9-Vl)/Vl -(vt-1.656)/vl =0.1274-\177v\177 =1.898.

Similarly, the specific voIumes at the other test points can be calculated. Using Equation
4.29 for the initial states we construct two equations with the two unknowns X and N:
1.840 = N - 2, In
200,0, 1.898 = N - 2,In 150.0. Solving for X and N: X = 0.2, N - 2.9.
The equation of the NCL is: v = 2.9 - 0.2In p',
Both final states are locatedon the CSL, hence:

300 0.6

\177'200 o 0.4
Undrained

100 0.2

Drained

0 0
0 100 200 300 0 0.2 0.4 0.6 0.8 1
p' (kPa) P'/P'c
(a) (b))))

Figure 4.37. Example 4.17.

 150 Soil Mechanics: Basic Co,cepts and Engineeritzg Applications

1.840- F - 0.2xln94.4 \177, F -

2.75,
The of the CSL on the is: v - 2.75 - 0.2 In
equation v-p' plane
(c) In the undrained testp' and q' values are divided
the by 200 (kPa) to normalise the
stress state. In the drained
test, the equation of the NCL (v -- 2,9 - 0.2 In p') is used to cal-
culate the preconsolidation pressures on the constant v curves (test points) which intersect
the drained stress path whi/e the sample progresses from the initial state towards the criti-
caI state. To normalise the stress state, all p' and q' values are divided by the preconsoIida-
tion pressure. The results are shoran in table below and the normalized stress paths are
shown in Figure 4.37(b).
Sample calculation for test point 5 in the undrained and drained conditions:

P'/P'C.
= 130.1/200.0=0.650,
C =821/2000 = 0.411.
\177v -AV/V 0=-(v 5-v\177)/v I =(1.898-v 5)/1.898=0.0499-+v5 =1'803,
v=2.9-O.21np\177 1.803--\177pc=241.tkPa.
P'/P'C =176.8/241.1=0.733.q'/PC =80.3/241.10.333.

Test Undrmned: \177a = 200 kPa Drained: \177 = 150 kPa

points q\" / 200 p' / 200 v P'c (kPa) q\" / P'c P' / P'c
0.000 I 000 .898 150.0 0.000 1.000
0.139 0900 .877 166.5 0.094 0.932
0.262 0800 .854 186\1778 0 181 0.863
0.366 0700 .830 2106 0 262 0.800
0.411 0 650 .803 241.1 0.333 0.733
0.449 0 599 .772 281.4 0.399 0.666
0.482 0 550 .738 333,6 0 448 0.599
0.508 0.500 .698 407.4 0.491 0.532
0.520 0 472 .656 502 7 0 518 0.471

4.6 PROBLEMS

4. i At a point 15 m below the ground surface, the relationship between the effective ver-
tical stress cy' a and the effective lateral stress o'r is: \177'\177= o'z(l -sinq\177'). If the water
table is 2 m below the ground surface, calculate the normaI and shear stresses on the
two perpendicular planes P and
Q where the angle cz t\177r the P plane is 45 \302\260
+ \177' / 2.

c =0, qa =40,9&). =l.TM\177m andp,ar=l.9\177M\177m.

Answers: o\177,
= 72.9 kPa. =
\"r\177, 38.0 kPa, OQ = 136.8 kPa, \"r0 = - 38.0 kPa
4.2 At a point within a soil mass, the effective lateral and shear stresses are 100kPa and
50 kPa respectively. Calculate the effective vertical stress to cause the failure of the

point, c' >- 0, \177' - 30 \302\260.

Answer: o\"z = 233.3 kPa
-\177.3 An clement of soil m xc piano is subjected to o'c, o',, and \"\177,c.Assuming that:

\"rxz
= 0.306 O'x. calculate the ratio o'x / cQ at failure, c' - 0, O' - 36 \302\260.

Answer: 0.271

4.4 The results of a direct shear test on a specimen of dry, sand are as follows: normal

stress - 96.6 kPa, shear s'\177ress at failure - 67,7 kPa. By means of a Mohr's circle of)))
 Shear Strength of Soils and Failure Criteria 151

stresses,find the magnitude and directions of the principal stresses acting on a soil
element within the zone of failure.
Answers: 0.I = 226.7kPa, 0\"3
= 61.4 kPa, = 62.5
\302\242t
\302\260
(from failure plane).
4.5 Data obtained from a drained triaxial test are as follows. Determine c\" and 0'-

Testno. \1773(kPa) \177- \1773(kPa)atpeak

1 50 191
2 100 226
3 150 261
Answers: c' = 60 kPa, d)' = 15\302\260
4.6 The results of three consolidated-undrained triaxial tests on identical specimens of a
particular soil are:

Testno. \1773(kPa) \177- \1773(kPa)atpeak u (kPa)atpeak

t 200 244 55
2 300 314 107

3 400 384 t59

Determine c' and d)'. What would be the expected pore pressure at failure for
= 100 kPa?
\1773

Answers: c' = 10kPa, q\177' u =

= 25 \302\260, 2.6 kPa
4.7 The results of drained and consolidated-undrained triaxial tests on two samples of a
normally consolidated clay are shown below:

test -
Type of the \177r3(kPa) \177rl (kPa)
\177r.\177 at peak

Drained 300 650

Consolidated-undrained 200 250

Determine: (a) \177' from the drained test, (b) q\177from the consolidated-undrained test,
(c) the pore pressure in the consolidated-undrained test at failure.
Answers: 31.3 22.6
\302\260, 84.6
\302\260, kPa

4.8 A soil has the following properties: n=0.38, Es--t0MPa, \177=0.3.The bulk

modulus pore of the water is 2200 MPa. Estimate the pore pressure coefficient B.
Answer: 0.9986
4.9 An unconfined compression test has given a UCS value of 126.6 kPa. The effective

shear strength parameters are: c' = 25 kPa, O\" = 30 \302\260.

Assuming the pore pressure pa-
A = - 0.09, calculatethe initial in the
rameter pore pressure sample.
Answer: - 8.6 kPa
4.10 The values of the critical state parameters for a particular type of clay are: N = 2.1,
\177=0.087, P = 2,05, M= 0.95. Two samples of this soil are consolidated under a

confining pressure of 300 kPa. One sample has been subjected to a drained triaxial
test whilst the second sample has been sheared in an undrained condition. Deter-)))
152 Soit Mechanics: Basic Concepts and Engineering Applications

mine: (a) the deviator stress at the critical state for both the drained and undrained
tests, (b) the pore pressure in the undrained test critical state, (c) the
at the volumet-
ric strain in the drained test when the sample approaches the critical state.
Answers: 417.0 kPa, 160.4 kPa, 184.6 kPa, 5.18%
4.11 In a drained triaxial test carried out on a sample of the clay of Problem 4.10, the

sample was first consolidated pressure of 400 kPa. It

under a confining was then
unloaded to 300 kPa and, after equilibrium was reached, it was sheared in a drained
condition. If the \177:value is 0.037, calculate the volumetric strain at failure.

Answer: 4.3%
4.12 The critical state parameters of a soil are: M = 0.857,\177.= 0.095, N= 2.1, F = 2.05,
\177:= 0.045. of this soil have been consolidated and unloaded to obtain an
Specimens
initial void ratio of 0.62. (a) If the
subjected to an undrained
specimenstriaxiat are
test, find the minimum overconsolidation ratio (OCR = m) above which the pore
pressure at the critical state becomes negative, (b) calculate the volumetric strains
for three specimens of OCR = 1, OCR = m
(as defined above) and OCR = 8 that are
subjected to drained tests.
= 5.15; for OCR = 1,ev = 5. l %
OCR OCR = 5.15,
Answers: (compression); for

\177v
= 0 %; for OCR = 8, \177v = 1.36% (expansion).

4.7 REFERENCES

Arthur, J. R. F., & Aysen, A. 1977. Ruptured sand sheared in plane slrain. Proc. intern, cotf SMFE
1 : 19-22. Tokyo.
Arthur, J. R. F., Dunstan, T., AI-Ani, Q. & Aysen, A. 1977. Plastic deformation in granular media.
Geotechnique 38( ): 1 140-144.
Arthur, I. R. F.. Dalili, A. & Dunstan, T. 1988.Discussion on the engineering application of direct
and simple shear testing. Geotechnique 27(i): 53-74.
Atkinson, J. H., & Bransby, P. L. 1978. The mechanics of soils.UK: McGraw-Hill.
Britto, A. & Gunn, M.J. 1987.Critical state soil mechanics viafinite elements. Ctlichester, UK: Ellis
Horwood.
Caner, J.P.,Booker, J.1., & Wroth, C.P. 1982.A critical state soil model for cyclic loadings. In G.N.
Pande & O.C.Zienkiewicz (eds.L Soil mechanics-transient and cyclic loads:219-252.New
York: John Wiley & Sons.
Jewell, R. A. 1989. Direct shear tests on sands. Geotechnique 39(2):309-322.
Powrie, W. 1997. Soil mechanics-concepts and applications. London: E& FN Spon.
m
Roscoe, K. H.1953. An apparatus for the application of simple shear to soil samples. Proc. 3 in-
tern. conf SMFE i : 186-191. Switzerland.
Roscoe, K.h. & Burland, J.B. 1968. On the generalized stress-strain behaviour of wet clay. In J.

Heyman & F.A. Leckie(eds.),Engineering plasticity: 535-639. Cambridge, UK: Cambridge Uni-

versity Press.
Roscoe, K.I\177. & Scl\177ofield, A.N. 1963. Mechanical behaviour of an wet clay. Proc. Euro-
idealized

pean conf on soil rnechanics and foundation engineering 1 : 47-54. Wiezbaden.

Terzaghi, K. & Peck, R. B.1948.Soilmechanics in engineering practice. New York: John Wiley &
Sons.
Wong, R. K. S. & Arthur, J. R. F. 1985. Induced and inherent anisotropy in sands. Geotechnique
35(4): 471-481.)))
CHAPTER 5

Stress Distribution and Settlementin Soils

5.i INTRODUCTION

This chapter describes the stress distribution and the calculation of settlement within an ideal-
ized elastic soil mass due to applied external and internal loading. The elastic properties in-
clude the Modulus of Elasticity Es and Poisson's ratio \177t.For soils where the compressibility
characteristicsare non-uniform and depend upon the state of the stress (e.g. sands),the con-
cept of a compressibility index my is introduced. The coefficient of (volume) compressibility is
defined as the ratio of the strain increment to the stress increment in a specified direction
caused by external or internal loading. The soil is assumed to be an ideal semi-infinite ho-

mogenous elastic material obeying Hooke's law, and the elastic stress-strain model shown in
Figure 4.4(a) is employed to determine the stress distribution within the soil. Furthermore, the
soil is assumed to be weightless which means the stress field resulting from the solution of the
equilibrium and compatibility conditions satisfies only the boundary conditions. Nevertheless,
with a known stress field the strains, and consequently the deformation of the soil layer and
surface settlement, can be evaluated. When the compressibility index concept is used, a mean
stress increment within a zone of influence is estimated by in-situ testing. Surface settlement
can be accurately predicted even though the stress distribution in the soil remains undefined.
In an ideal semi-infinite elastic soil the stress field under a specified loading is independent
of the Modulus of Elasticity, but does depends on Poisson'sratio. However, in most of the
loading types described in this chapter, the vertical stress (parallel to the load) is also inde-
pendent of Poisson's ratio. The magnitude of the Modulus of Elasticity affects the prediction
of elastic deformations and surface settlements. Moreover, the reliability of the predicted
stresses dependsupon the stress-strain model, and the accuracy of the etastic parameters. Un-
fortunately, the elastic parameters determined in the laboratory are not representative of field
conditions due to differences in the density, moisture content and stress history.
Stresses and settlements within a soil mass are caused by both external and internal loading.
External loading includes vertical loads applied on the ground surface or near the ground sur-
face. Internal loading is applied inside the soil mass away from the ground surface (e.g. piles)
and may include a vertical concentrated force or a distributed shear stress, or a combination of
both. In the case of internal loading, only axisymmetric conditions are consideredin this chap-
ter. Surface loads are transmitted through a footing, and the stress distribution will depend on
153)))
154 Soil Mechanics: Basic Concepts and Engineering Applications

the interaction model selected for the contact the soil and the
area between footing. In general,
the contact pressure between a footing and the somewhat indeterminate
soil is and depends on
the rigidity of the footing and the type of soil. If the thickness of the footing
satisfies a rigidity
criterion, then the contact pressure can be assumed uniform or linear, depending on the loca-
tion of the resultant of the applied forces. If, for a given plan size of footing, the thickness does
not satisfy the rigidity criterion, then the solution for the contact pressure is obtained by as-

suming the footing acts as an elastic beam resting on an elastic foundation. In this chapter, the
resulting stress field within a soil is determined assuming a certain stress distribution at the
boundary without taking account of the interaction model. Elastic settlements are predictedfor
two extreme conditions of footing stiffness viz., flexible and perfectly rigid.

5.2 FUNDAMENTAL EQUATIONS OF ELASTICITY

5.2.1 Stresses and equilibrium

In two-dimensional space, the state of stress is represented by the three stress components, \177x,
\177,and Zxz, as shown in Figure 5.1(a). Partial derivatives define the change in stress in the
specified direction. For convenience, the positive sense of the z-axis is selectedin the direction

of gravity. Equating the sum of the forces in the x and z directions to zero, two corresponding
differential equations of equilibrium for the element are obtained:

3\302\260'x &tvc = 3\302\260'---\177z

+ 0, + 3\"rx-\177-z= y (5.1)

where y is the unit weight of the soil.

(a)

(b)

Figure 5. I. (a) Two-dimensional stresses on an element, (b) two-dimensional displacements and strains.)))
 Stress Distribution and Settlement in Soils 155

A two-dimensional stress field describingthe three unknown components of stress at every

point in the soil cannot be created by considering only the two equations of equilibrium. That

is, whilst these equations are necessary,they are not sufficient, and a third condition is needed
to relate the state of stress to the elastic behaviour of the soil. Compatibility of deformations or
strains to keep the body intact condition and, in an elastic analysis, this condi-
creates the third
tion is based on Hooke's law the proportional relationship
that defines
between stresses and
strains. Furthermore, the stress field must satisfy the stress boundary conditions within the

equilibrium equations. In three-dimensional space there are six components of stress that sat-
isfy the three differential equations of equilibrium:
3% 3z\177
+ + + + =0' (5.2)
ax \177 a\177=0'\177,, \177 \177,\177, az \177x ay

It can be seen that at least three independent conditions of compatibility of deformations or

strains are neededto define a three-dimensional stress field.

5.2.2 Strains and compatibility

Figure 5.1(b) shows the deformation of a two-dimensional element in x, z coordinate system.

Two parameters u (in the x direction) and w (in the z direction) representthe displacement of
point o to a new position o'. The two sides of the element (oa and ob) undergo axial deforma-
\302\260
tion as well as rotation. These deformations, and the total change in theoriginal 90 angle, are
described by axial strains in the x and z directions (ex and ez) and the shear strain (Yzx):

au aw \177w
\177x \177z \302\245\177 (5.3)
=\177-x' ---\177-z' =\177+ az

The compatibility condition in two-dimensional space is expressed by:

a2Ex \177a2gz \1772\177 = 0
(5.4)

In three-dimensional space the following relationships between strains and displacements ae

additional to Equation 5.3:

3v av au aw av

where v is the displacement in the y direction. In an ideal elastic materi\177 Hooke's law relates
the axial strains to the normal stresses by the following linear equations:
1 l
ex =
x -\177y -\177Oz), ey -\177O z -\177Ox),
\177(O =\177(Oy
l
e. = \177(o\177
-
\177ox
-
\177o v) (5.6)

where Es is the Modulus of Elasticity and \177is Poisson's ratio. The values of Ev and \177depend

on the drainage conditions (undrained or drained) and are estimated from experimental stress-)))
156 Soil Mechanics: Basic Conceptsand Engineering Applications

strain relationships obtained in appropriate laboratory triaxial tests, or from in-situ loading
tests. It may be shown that the shear strains are proportional to shear stresses and are given by:

T'xy _ T'Y z _ ZzX = Es

Yxy Yyz
--'\177G (5.7)
=--\177-, \177-,Yzx -'-'\177 \177
2(l+g'-\"
where G is the shear modulus. The volumetric strain is approximated by the sum of the axial
strains:

AV - 1-2g
ev =--=\177x+eY (l\177x+l\177y +Oz) (5.8)
+eZV Es
In this equation the stress terms may be replaced by the incremental stresses as indicated in

Equation 4.21. In undrained loading no volume change occurs and therefore:

AV 1- 2g,
\177v- -0, _--\177,ox+\177y+\177z)=0, g=0.5 (5.9)
V

Substituting Equation 5.9 into 5.7 we obtain:

Es = 3G or Esu = 3G (5.10)
where Esu is the Modulus of Elasticity in undrained conditions. Since the shear stresses are re-
sisted solelyby the soil particles and their magnitudes are independent of the drainage condi-
tions (or pore pressures),it is reasonable to accept that, in an isotropic elastic soil, the shear
modulus for drained and undrained conditions are alike.

Example 5.1
A 200 mm cube of sand is subjected to
(\177y
=
0) loading with ox = 150kPa (com- plane strain
pression), oz = 800 kPa
(compression) The applied stresseshave caused 6.4 mm
and zxz = 0.

compression and 0.6 mm expansion at the z and x directions respectively. Determine: (a) E\177.
and g for the sand, (b) the normal stress in the x direction to recover the 0.6 mm expansion and
overall compressionin the z direction.
Solution:
(a) Substituting -- 0 in Equation 5.6: = + = + 150.0) = 950.0\177t.
\177y Oy
\177t(c\177
z Ox) \177t(800.0

Replacing oy in \177xand ez (Equation 5.6), we obtain two equations with the two unknowns Es
and \177t:

800.0- g(150.0 + 950\177t) =

(6.4 / 200.0) E s = 0.032Es,
-
150.0 \177t(800.0 + 950g)
= -(0.6/200.0)E
s = -O.003Es . Solving for Es and \177t:

= 0.2165. =
oy 950.0g 205.7
Es
= 22592 kPa = 22.6 MPa, g = kPa.

(b) Planestrain ( ex = ey = 0, ox --
conditions apply in both x and y directions
oy ):
= - - / E s = 0.0,
\177x (\177x \177t\177y 800.O\177t)
- - 800.Ox 0.2165 = 0.0 --\177Ox = - 221.0kPa.
Ox 0-2165C\177x oy
ez
= [800.0 0.2165 x (221.0 + 221.0)]/22592= 0.03117,
bz = 0.03117 x 200 mm = 6.23 mm.)))
 Stress Distribution and Settlement in Soils 157

5.3 STRESS DISTRIBUTION DUETO EXTERNAL AND INTERNAL LOADING

5.3.1 Typesof loading

The state of stress within an elastic soil can be studiedin two- and three-dimensional systems.

Stressesgenerated under the following loading conditions are examined in three-dimensional

space:
1. A concentrated vertical load applied at the ground surface.
2. Loaded areas using concentrated loads and the concept of superposition.
3. A circular loaded area and the construction of influence charts.
4. A rectangular loaded area.
5. A concentrated line load of finite length.
Two-dimensional states of stress are examined for the following loading conditions:
6. A concentrated line load of infinite length.
7. A uniformly or linearly loaded infinitely long strip.
In all cases the soil is assumed to be a semi-infinite elastic, homogeneous, isotropicand
weightless material (the weight of the soil does not affect the pattern of stress distribution).

5.3.2 Concentratedvertical load

More than a century ago Boussinesqgave solutions for the stress distribution within an ideal-
ized elastic material under a concentrated vertical load appliedto a horizontal ground surface,
Body forces were not included in the analysis, and the stress field described only the effect of
the external concentrated load. The stress field satisfied the three conditions of equilibrium,
compatibility and the stress boundary conditions. Although independent of the Modulus of
Elasticity it does depend on the magnitude of Poisson's ratio. Boussinesq's stresscomponents
may be presented in variety of coordinate systems (Jumikis, 1969;Poulos & Davis 1974).

Figure 5.2. Stress components at a point due to a concentratedload.

L)))
158 SoilMechanics:BasicConcepts and Engineering Applications

However only the cylindrical coordinate system (z, r) and the combination of cylindrical and

polar coordinate systems (t3, z) have been considered here. Equations 5.11-5.14 define the
normal and shear stresses in the cylindrical coordinate system shown in Figure 5.2.

- Q 3z 3
o
z 2 (5.11)
2\177z
(r + z2)5/2

3r2z l-2g
=_\177_Q
Or [
5/2 1/2] (5.i2)
2\177Z (r2 + z2) r2 + z2 + z(r2 + z2)

Q z 1
\302\260t=---(1-2g)[2+ 1/2] (5.i3)
z2)3/2 r
2
2\177: (r +z 2+z(r 2+z2)
2
Q
3r z
T'rz
=
2 5/2 (5.14)
271;
(r + Z2)

Equations 5.15-5.18 represent the same stresses in terms of the angle 13 and depth z.

oz =
(3cos50) (5.15)
,_\177Q2
2\177tz

Or= Q_
z
[3sin213cos313 (l-2g)c\302\260s2131 (5.16)
2\177zz 1+ cos 13

2
ot Q
2g)(cos3 0-
cOS
--0 )
2 (1- 1 +cosO
(5.17)
2/ZZ

- Q
Y'rz (3sinOcos40) (5.18)
2KZ2
A common representation of the vertical stress component in terms of the dimensionless pa-
rameter r / z and an influence factor lq is:

_ Q 3 - Q 3
o
z lq, lq- (5.19)
z2 2\177z[(r/z)2 +1)]5/2 z2 2N(r/z)2 +1)]5/2
Generalpatterns of vertical stress component at sections
of the distribution for r \177-0, constant
z, and constant r are shown
5.3(a), 5.3(b) and 5.3(c) in Figures
respectively. When r is con-
stant, it is a requirement of the boundary conditions that the vertical stress component is zero
on the ground surface, increases to a maximum at a specified depth and decreases as z in-
creases. It may be shown that the r / z ratio is a constant for the maximum stress point regard-
less of the distance from the applied concentrated load.
Equations 5.16 and 5.17 represent the horizontal stresses in the radial and tangential direc-
tions. Depending on the limiting value of the angle 13 these stresses are either compressive or
tensile.For the horizontal radial stress or, the limiting value depends on the magnitude of
Poisson's ratio. For example,for \177t= 0.2 it may be shown that the limiting value of 13 is 79.9 \302\260.
For angles less than 79.9 \302\260,
Or is compressive. Or is zero whilst for 0 > 79.9\302\260,
When 13 = 79.9 \302\260,
or becomes tensile. Consequently, for \177t= 0.2 the horizontal radial stress outside a vertical)))
 Stress Distribution and Settlement in Soils 159

\177z
r=constant\177i '\177 I -Y
\177-\".\177t , !
\177 Z= c
F\177_\177
r
onstan__.t\177.l__

(a) (b) (c)

5.3. S[ress dis[ribufioa uader a coacemra[ed load.

\302\260
angle of 2 x 79.9 = 159.8
\302\260
cone with a central is tensile. As the value of \177tincreasesthe corre-
sponding central angle of the vertical cone increasesuntil, at =
\177t 0.5, the central angle be-
comes180\302\260.
Similarly, it may be shown that the horizontal tangential stress ot for all values of
\302\260 \302\260
\177t< 0.5 is tensile inside a vertical cone with a central angle of 2 x 51.83 =103.7 and com-

pressive outside this cone,

The majority of in the z direction. Early
the deformation attempts to compare the
occurs
calculated and measured were described in the technical reports of the Corps
vertical stresses
of Engineers in the USA (Wu, 1966). These studies show that the Boussinesq's stress distribu-
tion may still be applied with reasonable accuracy when estimating the stress conditions in a
soil near the loading area.

Example 5.2

Under vertical load of 100kN, determine:

a concentrated

(a) the distribution of oz on horizontal planes at depths of 1 m, 2 m, and 3 m below the ground
surface - vary r / z from 0.0 to 3.0 in 0.25 increments,
(b) the distribution of cyz on the vertical planes at 1 m, 2 m, and 3 m from the applied concen-
trated load and the position of the maximum vertical stress on these planes- vary z in 0.5 m
increments.
Solution:
Using Equation 5.19 the vertical stress component for both cases are calculated and tabulated

in the following tables and are shown in Figure 5.4.

(kPa) cYz (kPa)

r/z ch
r/ z
z=l.Om z=2.0m z=3.0m z=l.Om z=2.0m z=3.0m
0.00 47.75 11.94 5.30 1.75 1.43 0.36 0.16
0.25 41.03 10.26 4.56 2.00 0.85 0.21 0.09
0.50 27.33 6.83 3.04 2.25 0.53 0.13 0.06
0.75 15.64 3.91 1.74 2.50 0.34 0.08 0.04
1.00 8.44 2. ll 0.94 2.75 0.22 0.05 0.02
1.25 4.54 1.13 0.50 3.00 0.15 0.04 0.0l
1.50 2.50 0.63 0.28)))
160 Soil Mechanics.\" Basic Concepts and Engineering Applications

r= l.Om r= 2.0m r=3.0 m

z (m)
r/ z cy z (kPa) r/ z cYz (kPa) r/ z % (kPa)

0.0 0.00 0.00 0.00

0.5 2.000 3.42 4.000 0.16 6.000 0.02
1.0 1.000 8.44 2.000 0.85 3.000 0.15
1.5 0.667 8.46 1.333 1.65 2.000 0.38
2.0 0.500 6.83 1.000 2.1l 1.500 0.63
2.5 0.400 5.27 0.800 2.22 1.200 0.82
3.0 0.333 4.08 0.667 2.11 1.000 0.94
3.5 0.286 3.20 0.571 1.92 0.857 0.98
4.0 0.250 2.56 0.500 1.71 0.750 0.97
4.5 0.222 2.09 0.444 1.50 0.667 0.94
5.0 0.200 1.73 0.400 1.32 0.600 0.88
5.5 0.182 1.45 0.364 1.16 0.545 0.82
6.0 0.167 1.24 0.333 1.02 0.500 0.76

In order to locate the depth corresponding to the maximum vertical stress component, we set
the derivative of the vertical stress (Equation 5.11) to zero:
3
Oo z _ 3Q z
0__[
5/2]=O--'3r2-2z2
=O-\177r/z=\177/3-\177O=tan-I 2\177f\177=39\"23\302\260\"
Oz 2\177zOz (r2 +z2)
= 0.47746/[(\177-\177)2 + 5/2 = 0.13314.
z = 1.2247r, from Equation 5.19: 1]
lq
2
For r = 1.0 m, z = 1.225m =
\302\260z [100.0/(1.225) ] x 0.13314 = 8.9kPa.
2
For r = 2.0 m, z = 2.449m o
z
= [100.0/(2.449) ]xO. 13314= 2.2kPa.
For r = 3.0 m, z = 3.674m = [100.0/(3.674) 2 ]xO. 13314= 1.0 kPa.
%

o\" (kPa)
z
0 2 4 6 8 10
6O 0
5O 1
2
\177 40

\177 30
N 4
20
5
10 6 .111r=am
7
0 0.5 1 1.5 2 2.5 3
(b))))

r (m)

(a)

Figure 5.4. Example 5.2.

 Stress Distribution and Settlement in Soils 161

= 400 kN i = 400 kN \177-

Q1 Q2 40
'
\1773o
,

\177 \342\200\242 \177 10

\177 \177 2.0m\177

2 4 6 8 10 12 14 16
(a)
Distance from point A (m)

(b)
Figure 5.5. Example 5.3.

Example 5.3
Referring to Figure 5.5(a), calculate the distribution of the vertical stress component from

point A to point B at a depth of 2 m. Specify the vertical stresses at I m intervals.

Solution:
For multiple concentrated loads, superposition is used to evaluate the stresses at any point
within the soil. Due to symmetrical loading, we will consider only the left-hand half.The re-
suits are presentedin the table below and illustrated in Figure 5.5(b).

Q\177 Q2
Points Total Cyz (kPa)
r (m) r/z cYz (kPa) r (m) r/z cy z (kPa)

A 5.0 2.50 0.34 10.0 5.00 0.3

1 4.0 2.00 0.85 9.0 4.50 0.8
2 3.0 1.50 2.51 8.0 4.00 2.5
3 2.0 1.00 8.44 7.0 3,50 8.4
4 1.0 0.50 27.33 6.0 3.00 0.15 27.5
5 0.0 0,00 47.75 5.0 2\17750 0.34 48.1
6 1.0 0.50 27.33 4.0 2.00 0.85 28.2
7 2.0 1.0 8.44 3,0 1.50 2.5 l 10,9
8 2.5 1.25 4.54 2.5 1.25 4,54 9.1

Example 5.4

Referring to Figure 5.6, calculate the normal and shear at point A with R = 3 m on a
stresses
plane that passes through the point of application of and makes an angle of 55 \302\260
the load below
the ground surface. Determine the magnitudes and directions of the principal stresses at this
point. Repeat the calculation for another point on this plane with R = 1.5 m. Assume bt = 0.3.

Solution:
z=R cos 0 = - 55.0
3.0 cos(90.0 \302\260 =
\302\260)2.457 m, using Equation 5.15:
5 =
c\177 = (400.0/2/tx2.4572)(3cos 35.0 \302\260)11.7 kPa.)))
z
162 Soil Mechanics.\" Basic Concepts and Engineering Applications

Q = 400 kN

o
\17755

= 3.0 rn
',,,R

\"
.\177- \177rz

Figure 5.6. Example 5.4.

\302\260
Substituting
= 0.3,
\177t z = 2.457 m, and 0 = 35 into Equations 5.16 and 5.18 we obtain:
err
= 4.2 kPa and Zrz
= 8.2 kPa. From Equations 4.1 and 4.2:
cy =
(11.7 + 4.2)/2 + (11.7 - 4.2)/2\303\227
cos(-2 55.0
\303\227 + 8.2
\302\260) s in(-2 55.0
\303\227 =
\302\260) 1.0 kPa.
=
\177: (11.7
- 4.2)/2 \303\227
sin(-2\303\227 55.0)
-
8.2cos(-2 \303\22755.0)
= -0.7 kPa.
0 = 1/2\303\227tan-l[2\303\2278.2/(l 1.7 -4.2)] = 32.7 \302\260.

Using Equations 4.3, 4.4 and 4.5: s = (or1 + cy 3 ) / 2 = (or z

+ Or ) ]2 = (i 1.7+ 4.2)/ 2 - 7.9 kPa,

2+
t = \177/[(11.7 -
4.2) / 2] (8.2)2 = 9.0kPa,
or3= s- t = 7.9-9.0 =-1.! k_Pa, CYl=S + t- 7.9 + 9.0 - k_Pa. 16.9
At R =1.5 m: z = Rcos0= 1.5cos35.0 z = 46.6 kPa Or = 16.6kPa vrz = 32.6 kPa.
\302\260
= 1.229 mcy

be seen that the stress components have by the ratto of (3.0/ 1.5) 4, there-
It can tncreased =

fore the direction of the principal stresses remains constant:

s- 31.6
kPa, t = 35.9 kPa, or3 = 31.6 - 35.9 = - 4.3kPa, or1
= 31.6 + 35.9 = 67.5kPa.

5.3.3 Loaded areas using superposition of concentrated loads

Stressesat a beneath a loaded area is found
specified depth by application of that developed
previously for load.
a concentrated
Mathematically this is possible by integration of the stress
equations within the loaded area. Alternatively, the loaded area may be sub divided into many
small areas (Figure 5.7) and the distributed load q within each small area replaced by a con-
centrated load acting at the centroid of the small area. Superposition of the effects of the finite
number of concentratedloads at the point of interest yields the desired stress components.

Example 5.5
The footing shown in Figure 5.8(a) is subjectedto a uniform load of 300 kPa. Calculatethe
vertical stress component at 2 m below point A, using the concept of superposition of concen-
trated loads.)))
 Stress Distribution and Settlement in Soils 163

Figure 5.7. Superposition of concentrated loads,

Solution:
Since the footing is symmetrical, the effect of the right-hand half can be calculated and then
doubled. The right-hand half is divided into 10 squares of l m by i m and the concentrated
load in each square is 300 kN (Figure 5.8(b)). Distances (r) between the centres of each square
and point A are calculated to establish r / z ratios.The results are shown in the table below.

Point r (m) r/z lq cy z (kPa) Point r (m) r/z lq

cYz (kPa)

1 3.535 1.7677 0.0138 1.04 6 1.58t 0.7906 0.1418 10.64

2 3.808 1.9039 0.0104 0.78 7 2.121 1.0607 0.0725 5.44
3 2.549 1.2747 0.0428 3.2l 8 2.915 1.4577 0.0276 2.07
4 2.915 1.4577 0.0276 2.07 9 0.707 0.3535 0,3557 26.68
5 3.535 1.7677 0.0138 1,04 10 1.581 0.7905 0.1418 10.64
Total %due to right-hand half= 63,6 kPa, total cy\177
= 63,6 x2 = 127,2kPa.

6.0 m

m .......... ......... _-\177

4.0 k--\177---l---t
L7_ l

(a) (b))))

Figure 5.8. Example 5.5.

i64 Soil Mechanics: Basic Concepts and Engineering Applications

5.3.4 Stresses under a uniformly loaded circular area: influence charts

Considera circular area on the horizontal ground surface with a uniform vertical loading of q
(Figure 5.9(a)).Vertical stress at a specified depth z maybe calculated mathematically by in-
tegration of Equations 5.i 1 or 5.15 within the circular loaded area, and the difficulty in this
process dependsupon the plan position of the complete solution for
point of interest. The
stresses, and deflections (or settlements) of the ground
strains surface is given by Ahlvin &
Ulery (1962). If the point is located on a vertical axis passing through the centre of the circle,
then the integration process for the vertical stress component is straightforward and yields:

- 3q 2n R c rdr
a z (5.20)
z j'\302\243
d0J'\177 2
2\177z [(r/ z) +115/2

Itis assumed that an elemental area dA defined by r x dr x dO (Figure 5.9(a)) is subjected to a

concentrated load of q x dA. A similar integration may be carried out for the stress component
ar. The results of the both integrations are presentedin Equations 5.21 and 5.22. Due to sym-
metry, the shear stresses on this axis are zero.

- 1
=
o\"
z q {1 2
} (5.21)
[(Rc / z) + 1]3/2
2(1+ pt) 1
\177r
r =a t +2\177t
2+
-\177
2+ } (5.22)
---\177{1
[(Rc / z) 111/2 [(Rc / z) 11312.
where RC is of the circular loaded area. Figure 5.9(b) shows an element of area abcd
the radius
defined by Rt and a central
and R2 angle of 2\177 / n. The integer n is selected as 4, 8, 12, 16,20
etc to ensure symmetry about the centre. It is assumed that this element is within a circular
area that has a very long radius approaching infinity. The vertical stress component calculated
from Equation 5.21 due to the loaded circle approaches the applied uniform vertical load q.
The vertical stress component produced by area abcd at depth z on the central axis, is obtained
by an integration similar to that shown in Equation 5.21:

rdrdc\177
,q

A (b)

O\"
t z

(a)

Figure 5.9. (a) Vertical stress under uniforrrfly loaded circular area, (b) the basics of influence chart.)))
 Stress Distribution and Settlement #\177Soils 165

1
(abed)
c\177.
\177
= q--(
312 3/2} (5.23)
n
[(RI/Z)2+I] [(R2/z)2+t]
Let us assume the area between two circles of radii RI and R2 produces a fraction m of the

contact pressure q (at depth z). The vertical stress due to the element abed is therefore:
\177 (abed) = mq / n (5.24)
z

Equating Equations 5.23 and 5.24 we obtain:

1 1
=m (5.25)
[(RI/z)2 [(Re/z)2 + 113/2
+ 113/2

Consider the case where m = 0.1 and Rt = 0.0. From Equation 5.25 R2 equals 0.26975z. Sub-

stituting this value of Rz as the internal radius R1 in Equation 5.25 we find the radius of an-
other external circle as 0.40050z.It is seen that a circular loaded area of radius 0.26975z pro-
duces a vertical stress of 0.1qat depth z on the central axis, while a circular area with radius of
0.40050z produces 0.2q; the difference between these is 0. lq. Similar calcuIations can be car-
ried out to produce the results presented in Table 5.1. If z is assumed to have a unit length then
Table 5.1 may be presented graphically by selecting a reasonable scale for z and a value for
n = 20 as shown in Figure 5.10. The total number of elements is n / m = 200 and each element
produces a vertical stress equal to (m / n) q 0.005q,where = lq
=- m / n = 0.005 is called the

influence factor. The resulting graph, called Newmark's influence chart (Newmark, 1942),is
used to estimate the verticaI stress component under a loaded irregular area. Initially a plan of
the footing is drawn assuming the depth z of the point of interest is equal to the scale of the
chart. The plan is placed on Newmark's chart so that the point (under which the vertical stress
component is required) is located at the central point of the chart. The vertical stress compo-
nent due to a unit contact pressure is equal to the number of elements within the plan multi-

plied by the influence factor, or

\177:
= (number of elements covered) x lq x q (5.26)
Table 5.1.Construction of Newrnark's influence chart wi\177h rn = 0.10.
R/z 0.00000 0.26975 0.40050 0.5181t 0.63696 0.76642
\177z/q 1.0 0.1 0.2 0,3 0.4 0\1775

R/z 0.91761 1.10970 1.38709 1.90829

\177rz/q 0.6 0.7 0.8 0.9 1.0

Example 5.6
Re-do Example 5.5 using the influence chart shown in Figure 5.10.
Solution:
A plan is drawn by assuming that the depth under point A is equal to the length of the scale in

Figure 5,10. The re-scaled figure is adjusted on the influence chart as shown. The best esti-
mate for the number of elements in one-half of the area is 42.5 Therefore:
\177
= 0.005 300.0\303\22742.5
\303\227 2 =
\303\227 127.5 kPa.)))
166 Soil Mechanics: Basic Concepts and Engineering Applications

Scale of depth z
I I
Influence factor = 0.005

Fi\177mare 5.10. Newmark's influence chart.

influence factor = 0.005)))

Figure 5.11. Example 5.6.

 Stress Distribution and Settlement in Soils 167

0,25 10.0
2.0
1.2
1.0
0.9
0.8
0.7

0.6
0.5

0.4

0.:3

0.2

0.1

0.1 1 10
i\"n

(b)

Figure 5.12. Vertical stress under a comer of a uniformly loaded rectangular area.

5.3.5 Vertical stress under a corner of a .,,iyor,,,fy loaded rectangular area

Fadum (1948) introduced influence factors (based on Newmark's mathematical approach) for
the computation of the vertical stress component at a specified depth under a corner of a rec-
tangular loaded area. The mathematical work is based on the integration of Equations 5,I 1
over a rectangular area. Two dimensionless paratneters rn and n are defined as:

re=L/z, n= B/z (5.27)

where L and B are the plan dimensions of the loaded rectangle,as shown in Figure 5,12(a).
The vertical stress component is computed from:
= (5,28)
\177Jz q\303\227lq

where Iq is the influence factor, a ditnensionless parameter defined by:

I . 2mn4rn2+n2+l m2+n2+2
lq
......... 2 2 2 2 2
+ tan
-\177 2mn@n2+t\1772+l
2 2 )
(5.29)
47r (m + tt + 1 + ttt2n m +n + l r,, + ,,= + t - rr,2r,

For negative values of tan -I, r\177should be added to

the calculated angle measured in radians.
Figure 5.12(b) shows the graphical presentation of Equation 5.29 where m and n are inter-

changeable dimensionlessparameters defined by Equations 5.27. The practical significance of)))

168 Soil Mechanics: Basic Concepts and Engineering Applications

this method is in its application to loaded areas of irregular geometry, The point at which the
vertical stress component is required may be located within or outside the plan projection of
the loaded area. Superposition may be applied if the loaded area is divided into a number of
rectangles having one common corner. If the point of interest is outside the loaded area then a
few of the rectangles that share the common comermay cover areas that are not loaded. In this
case it is convenient to assume that the rectangles that cover the unloaded areas are subjected
to a negative contact pressure,

Example 5.7
The uniform contact pressure under a rectangular footing of 6 m by 5 m is 200 kPa. Compute
the vertical stress component under points A and B (Figure 5.I3) at a depth of 2 m,
Solution:
The loaded area is divided into the 4 rectangles shown. Note that all rectangles share a com-
mon corner of either A or B. Results of the calculations are tabulated below.

Rectangle L (m) B (m) m n lq \177z(kPa)

Aebh 4.0 2.0 2.0 1.0 0.I999 39.98

Ahcf 2.0 1.0 1,0 0.5 0.I202 24.04

Afdg 4.0 1.0 2.0 0.5 0.1349 26.98

Agae 4.0 4.0 2.0 2.0 0.2325 46.50

Total verticaf stress component al A: 137.5 kPa.

Rectangle L (m) B (m) m n lq \177: (kPa)

Bebj 7.0 2.0 3.5 1.0 0.2039 40,78

Bfcj 2.0 2.0 1.0 1.0 -0.1752 -35.04
Bidf 4.0 2.0 2.0 1.0 -0.1999 -39.98
Biae 7.0 4.0 3.5 2.0 0.2387 47.74
Total vertical stress component at B: 13.5 kPa.

6.0 m
_.\177.
I,
e b
\177a

5.0 m

7.0m

!
-ILd \177 c

4.0m ,
!
i L ............ J ....... j
B

Figure 5. I3, ExampIes 5.7.)))

 Stress Distribution and Settlement in Soils 169

6.0 m

la b

/ i

Figure 5.14. Example 5.8.

Example 5.8
Re-work Example 5.5 using Equation 5.29 or Figure 5.12(b).
Solution:
Sincethe footing is symmetrical about Aa (Figure 5.14),computations are carried out on one-
half of the figure, the results of which are tabulated below.

Rectangle L (m) B (m) m n lq \177z(kPa)

Aabc 4.0 2.0 2.0 1,0 O. 1999 59.97

Adef 3.0 3.0 1.5 1.5 0.2167 65.0I
Adgc 3.0 2.0 1.5 1.0 -0.1936 -58.08
Ahif 3.0 1.0 1.5 0.5 -0.1313 -39.39
Ahjc 2.0 1.0 1.0 0.5 0.1202 36.06
Total \177rzat A
= 2(59.97 + 65.01 - 58.08 - 3%39 + 36.06)-- 127.I kPa.

5,3.6 Concentrated line toad offinite length

Figure 5.15 shows a vertical line load of finite length L applied to a horizontal ground surface
whose intensity q is measured by force per unit length (e.g. kN/m). For the arrangement
shown, the vertical stress component is defined by:

\177Jz =(q/
+ (5.30)
z)(lq, y lq,L_y)
where Iq,y and
lq,L_y
are influence factors given by:

1 3n n
= (- ,)3] (5.31)
lq'y'1q'L-Y 2 2 +n 2 2 2
2\177(m2 +1) [4m +l \177/m +n +1

in which the dimensionless parameters m and n are:

m= xt z, n =
y/ z (for lqy), n = (L-
y)/ z (for l q, L_y) (5.32))))
170 Soil Mechanics: BasicConcepts and Engb2eering Appticat(ons

'
y

A
Y

Figure 5.15. Vertical stress due to a line load of finite length.

For values of y greater than L the principle of superpositioncan be applied:

\177_ =(q/ -l L) (5.33)

z)(Iq.y q.y_

Example 5.9
A 6 m long line load of intensity 100 kN/m is applied to the ground surface. Compute the ver-
tical stress component at points A, B, C, and D located on a line at a depth of 2 m and 2 m

horizontaIly from the line toad. YA

= 0,
YB
= 3 m, YC,
= 6 m, = 9
YZ\177 m.
Solution:
At point A:
- x / z = 2.012.0 = z = 0.0/ 2.0= 0.0for
m 1.0,n =
y / [q.y,
n = (L - y) / z
= 6.0 / 2,0 = 3.0for lq.L -y: Iq,y
= 0.0,
lq\177L -y
= 0,07852,
=
(100,0/2.0)(0.0 +0.07852) = 3.9 kPa,
\177:A

At point B:
m= 2,0 / 2.0 = 1.0,n = 3.0 / 2.0 = 1.5for lq,y,
n = 3.0 / 2.0 = 1.5for lq,L -y:
= L -y
= 0.07 152, CrcB =
(100,0/2.0)(0.07152 + 0.07152) = 7.1 kPa.
lq.y lq,

Due to symmetry, the vertical stress at point C equals the vertical stress at point A (which is
3.9 klPa). At point D:
m--2.0/2.0 =l.0, n=9.0/2.0=4.5forlq.y, andn=O'-L)lz=3.0/2.0=l,5forlq.L-y:
lq.y
= 0.07933,
lq, L y
=
0.07152.6:D
= (100\1770/2.0)(0.07933-0.07152) = 0.4 klPa.

Example 5.10

Re-work Example 5.5 using the concepts of a line load of finite length and superposition. As-
sume that tl\177e plan view of the loaded areais divided into horizontal strips 1 m wide and that

the line load is applied on the centre line of each strip,

Solution:)))
 Stress Distribution and Settlement in Soils 171

The equivalent line loads are idealized by four horizontal strips (Figure 5,8) as follows:
Strip 1 includes squares 1 and 2; strip 2 includes squares 3, 4, and 5; strip 3 includes squares 6,
7, and 8; and strip 4 includes squares9 and 10. The results of computations are tabulated be-
low. A sample calculation for strip 3 is as follows:
For m = x / z = 1.5/ 2.0= 0.75, n = y / z = 3.0 / 2.0 = 1.5.For
[q,y:
- n = 3.0 ! [q.L y: 2.0= 1.5.
lq.y
=
[q,C -y
= 0.1207, q = 300.0kPa x 1.0 m (width of the strip)
= 300.0 kN/m.
0z = 36,21kPa.
Iq, L_y) =(300.(?/2.0)(0.1207+0.1207)
= +
(q/z)(lq, y

Strip no, m
t'q, y to, \",, _/. o. (kPa)
for'. lq, y for: lq. y
_ L

1 1.00 1.00 1.75 0.0120 0.0120 3.60

\"\177
1.50 1.50 1.25 0.0420 0.0420 12.60
3 1.50 1.50 0.75 0.1207 0.1207 36.21
4 1.00 1.00 0.25 0.2469 0.2469 74.07
Total oz = 3.60+ 12.60+36.21+ 74.07 \177-126.5 kPa.

5.3.7 Concentrated line toad of infinite length

The analysis of a line load of infinite length is normally considered in two-dimensional space,
and stresses are defined in accordance with the sign convention of Figure 5.16:

2 2q x2z
\302\260x
= cos2 0sin 0 = (5.34)
2
7rz2q ,rr,(x + \177-2)2
3
2q z
cr = 2qcos4 0 (5,35)
z
7\177Z \177(x2 + Z2)2
2
2q xZ
= 2q cos3 0sin 0 (5.36)
\"r.t z 2
r\177z \177(x + z2)2

Figure 5.16. Stresses at a point due to a line load of infinite length.)))

 I72 Soil Mechanics: Basic Concepts and Engineerhzg Applications

Example 5. l 1

Determine the stress components, principal stresses and their directions under a line toad of in-
finite length at points with x / z ratio of 0.6.
Solution:

Substituting x = 0.6 z into Equations 5.34, 5,35 and 5.36 we obtain:

Crx= O.124q/z, or.= 0.344q/z,'rxz= 0.207 q/z.
The position of the centre and the radius of the Mottr's circle are given by:
s = (\177z
+ crx)/2 = (0.344+ O.124)/2(q/z) = 0.234q/z,
2 2+
0,207 2 q / z = 0.234q/ z,
- cr 2 -
= 2] + Z.:x =
t
\177/[(\177: x) / \177/[(0.344 0.124) / 2]
cr
3=s-r=0,cr l=s+t=0.468q/z.
The direction of the major principal stress plane is calculated from Equation 4.6. Substituting
Equations 5.34, 5.35 and 5.36 into Equation 4.6:
2
2x \177
x. 2xz 2\303\2270.6z
tan20 t
= =--= = 1.875, from which tan0t =0.6 =tan0.
2 2 2 2
\177_ -o\"
x z -x z -(0.6z)
This means that at any point on a plane with x / z = constant, the angle between the major
principal plane and the x-axis is 0.

5.3.8 Uniformly loaded infinite strip

Stress components under a uniformly loaded strip of infinite length are calculated by integrat-
ing an infinite line load across the width of the strip. Such a loadedstrip is shown in Figure
5.17, Any point under the strip is represented either by its coordinates in the xz plane whose
origin is at the centre of its width, or by the angles \177xand 13. Equations 5.37 to 5.42 give the

Figure 5. [7. Stress components under uniformly loaded infinite strip,)))

 Stress Distribution and Settlement in Soils I73

stress components on any plane that is perpendicular to both the horizontal ground surface and
the infinite dimension of the strip. As for the line load, none of the stress depend components
on the elastic properties of the material. The vertical stress component may also be approxi-
mated using the Newmark's influence chart (Figure 5.12) or Equations 5.28 and 5.29 where ei-
ther m or n approacbes infinity, Tbe stresses are given by:

1 x+b 1 x-b 2bz(z2 -x2 +b2)

= q[tan_ tan_ ] (5.37)
crx 2 +x 2 2 2
\177; z z (z -b2) +4b2z

2 2 +b
-1 x+b
_tan_ 1 x-b
--+ 2bz(z -x 2)
or_
\177=\177[tan 2 .] (5.38)
rl; z z (z + xZ - be)2 + 4b2z 2

2
q 4bxz
\177: = [ ] (5.39)
:z 2
+4b2z 2
x\177 2
7z
(z +x -b2)

ox = q [c\177
- sin c\177
cos(c\177 + 213) ] / \177 (5.40)

o = q [o\177+ sin o\177cos(o\177 + 21\177)]/\177 (5.41)

z

= q [sin \177sin(c\177 + 2\177)]/\177 (5.42)

Zxz

Angles and
\177x 13 are defined by:
- -
z] -
1 -\177 - 1 -
c\177= tan [(x + b) / b) / z ] tan [(x b) /
(5.43) z ], 13
= tan [(x

Curvesrepresenting magnitude of a stress component (isobar)may

a constant be constructed
by equating each stress component to a specifiedconstant value. This is of particular interest,
as the volume of soil subjectedto deformation (as a result of the stress distribution) may be
evaluated by drawing an isobar representing a fraction of the applied load. The construction of
an isobar is fully explained in Example 5.13.

Example 5.12
An infinite strip of width 20 m is subjectedto a uniform load of 100 kPa. Compute the vertical
stress component at a depth of 8 m at points A (x = 0), B (x = 10m\177 and C (x = 18 m).
Solution:
At point A: o\177
a
= tan -\177
(10.0/8.0)
- tan-t / z = 51.34\177+
(\17710.0/8) 51.34
\177
= 102.68 \177,

13A
= tan-I (-10.0/8.0) = -51,34\302\260,
\302\260
C\177zA
=
100.0[102.68(\177/180.0) + sin 102.68\302\260cos(102.68 + 2\303\227-51,34\302\260)]/rt = 88.1 kPa.
At point B: \177n
=
tan-l(20.0/8-0) =68'20\302\260, [\177B =0.0,
\302\260
c\177z\177
=
100.0168.20(\177/180.0) + sin 68.20\302\260cos(68.20
=
+ 2\303\2270,0\302\260)]/7z 48.9 kPa.
At point C: = tan-I (28.0/8.0) - tan
-1
(8.0/8.0) = 74.05 -
\302\260
45.0 \302\260
= 29.05 \302\260,
\177c
1 \302\260
13 C
= tan (8.0/8.0)
= 45,0 ,
\302\260
= t
\177:\302\242 00.0129.05(\177/180.0)
+ sin 29,05\302\260cos(29.05 + 2 \303\227 =
45.0\302\260\177]/\177z8,6 kPa,)))
174 Soil Mechanics: Basic Concepts and Engineering Applications

Example 5.13

Under a uniformly loaded strip construct an isobar for \177z

=
0.2q.
Solution:

Step procedure: (a) obtain expressions for x and z in terms of the angles \177and 13, (b) select a
numerical value for \177and substitute this into Equation 5.41 and find [\177,(c) substitute \177,and [3
into the equations for x and z and calculate the x and z of the corresponding angles, (d) repeat
steps(b) and (c), From the geometry of Figure 5.17:
tan(a +[\177) =- (x+b)/z, tan\1773
=
(x-b)/z, solving for x aa\177d z:
x = b[tan(o\177 + 13) + tan [3]/[tan(o: + [3)- tan 13],z = (x + b) / tan(o\177 + [3).
For ct, = 30 \302\260
or \177/ 6 = 0.523598 radian:

cr: = q [a + sin ct, cos(ct, + 2[\177)]/r\177

=
0.2q, a + sin c\177cos(\177 + 2[3) = \177t=
0.2\303\227 0.628318,
\302\260 \302\260 \302\260 \302\260
0,523598 + sin 30.0 cos(30.0 + 213) = 0,6283 t 8, cos(30.0 + 2[3) = 0.209439 ,-\177[3 = 23.95

x = b [tan(30,0
\302\260
+ 23.95 + tan
\302\260) 23.95 \302\260]/[tan(30.0
\302\260
+ 23.95 - tan
\302\260) 23.95 \302\260
]
= i ,9556b,

x / 2b = 0.978, z = (1,9556b + b) l tan(30.0

\302\260
+ 23.95 =
\302\260)2.1513b ,-\177 z / 2b = 1,076.

Results of the computations (for the right half) are tabulated below and shown in Figure 5.18,

\302\260 \302\260 \302\260 \302\260

o\177 13 x/2b z/2b \177 [3 x/2b z/2b

60.0029.46 0.505 0,009 24.130 17,50 1.054 1.756

5&O0 29.30 0\177618 0.211 22.00 13.64 1\177011 2.107
38.00 27.62 0,81 I 0.594 20.00 7.63 0,844 2.567
30.00 23.95 0.978 1.076 19,00 2.65 0.632 2.852
28.00 22.35 1.017 1.257 18.50 -1.39 0.427 3.011
26.00 20.27 1.046 1.479 18.15 -9.075 0.000 3.130

1.5
0 x/2b

0.5

1.5
Oz= 0.2q \177'

2..5

3.5

Figure 5.18. Example 5.13.)))

 Stress Distribution and Settlement in Soils 175

5.3.9 linearly loaded infinite strip

Figure 5.19 shows a vertical distributed load applied to an infinitely Iot\177g strip and linearly in-
creasing across its width 2b. Using the sign convention shown, the stress components may be
obtained by integration of the load across the width of the strip. As before, all the stress com-
ponents are found to be independent of the elastic properties of the material.
x - R
2
= - L + sin
o,r In
\177 213)/2n (5
q(\177-o\177 b
= - sin
213)/ 27\177 (5.45)
q(\177ct

= q(l + cos 213 -

o\177)/ 2\177 (5.46)
\177

Angles c\177and 13 are defined by:

- 1 - - 2b) / z], -1 -
c\177= tan (x / z) tan -1[ (x 13
= tan [(x 2b) / z] (5.47)

Example 5.14
A load on an infinitely long strip increases linearly from zero to a maximum of 100 kPa across
its width of 8 m. Calculate the vertical stress components at a depth of 8 m at points:
A (x =0), B (x = 8
m) and C (x= 18 m).
Solution:

At point A:
c\177A
= tan -1 (0.0/8.0) = 45.0\302\260,
- tan-l[(0.0-8.0)/8.0] 13A
= tan-t [(0.0- 8.0) / 8.0]= -45.0
\302\260,

Figure 5.19. Stress components ut\177der a linearly loaded mfiniwly long strip.)))
176 Soil Mechanics: Basic Concepts and Engineering Applications

100.0
= = 15.9 k_Pa. At B:
\177-,.A [0,0-sin(2\303\227-45.0\302\260)] point
2n
-t
ct
B
= tan (8.0/8.0) - tan-\177[(8.0- 8.0)/8.0]
= 45.0 \302\260,
[\177B
= tan-l[(8.0
- 8.0)/8.0] = 0.0\302\260,
100,0 7\177
\302\260
(8.0 0
\303\22745 -0.0)
= 25.0 kPa.

= tan -1 = 14.70\302\260,
At point C: c\177
C (18.0/8.0)-tan-l[(18.0-8.0)/8.0]
\1773
C
= tan-l[(18.0- 8.0)/8.0] = 51.34\302\260,
100.0 18.0
qcc
= [ x 14.7\302\260 -
\177---\177--
sin(2x 5 t.34\302\260)]= 2.8 kPa.
2n 4.0 180\302\260

5.3.10 Dnbankment loading of infinite le\177gth

An earth embankment may be idealized as a combination of the uniformly and linearly loaded
infinite strips shown in Figure 5.20. The magnitude of the uniform load is ,/h, where y is the
unit weight of the earth embankment material of height h. The vertical stress component at
point A beneath the corner of the uniform load is the sum of Equations 5.41 and 5.45:

cr z =q[o: 2 +sinotzcoso: 2 +(a+b)\177lta-O.5sin2o;2]/zt,

o = q [o\1772 + (a + b)c\177 1 / a ]/ n (5.48)

z

The vertical stress component may be represented by the dimensionless parameters m = a/z
and n = b / z, obtained by substituting the magnitudes of tano:l and tahoe2 (calculated from the

geometry of the embankment) into Equation 5.48:

I m
i_tan_ln ) (5.49)
\177_=qlq, lq=l(m+\177-\177tzn

1'

O\"
z

Figure 5.20. Vertical stress under an embankment.)))

 Stress Distribution and Settlement in Soils 177

i
Centre line
10.0

Figure 5.2 l. Example 5.15.

If the point at which the vertical stress component is required is located within the linear or
uniform part, rather than the vertical edge of the emba,kment, superposition may be used.
Example 5.15
For the embankment shown in Figure 5,21, calculate the vertical stress component at points A
and B. Assume the unit weight of the embankment material is 20 kN/m 3.

Solution:
Vertical stress component at point A: Due to symmetry we need only to calculate the vertical
stress component due to the left half and double the result. For the left half:
m= a/z = 20.0 / 10,0 -- 2.0, n = b / z = 5.0/ 10.0= 0.5,using Equation 5.49:

1,2.0+ 0.5 2
0.5) = 0.4367'
_\177
tan + tan_l
IqA
=\177t\"\177-\177.0 1+0.52 +2x0.5
=
2x(20.0x10.0)\303\2270,4367 = 174.7 kPa.
cr,jt

Vertical stressat point B:

component
The section earth embankment is divided into the three areas cde, dflfi, and efg (Figure
of the

5.22(a)), where the area efg represents a negative linear loading. All three areas are analysed
using Equation 5,49 and the results are superimposed to give the vertical stress component at
point B, as shown in Figure 5.22(b). From the geometry of the earth embankment we find:
cd= 15.0tn, de= 0.5 x 15.0=7.5 m, ef = m, fg =5.0 10,0-7.5=2.5
m, flz= 15.0 m, The
results of the computations are summarised in the table below.

Area q
=y h (kPa)
\303\227 m n \177r
z (kPa)
lq

cde 20.0 x 7.5 = 150 1.5 0.0 0.3128 46.92

dflzi 20,0 x 10.0 = 200 2.0 1\1775 0.4853 97.06

efg 20.0 x -2.5
- -50 0.5 0.0 0.1476 -7.38
= 46.92 + 97.06 - 7.38 = I36.6
cy:\177 kPa.)))
I78 Soil Mechanics: Basic Concepts and Engineering Applicatiot\177s

i
Centre line
10.0 m

Bo m \302\260A

15.0 rn \177 10.0

rn
I

(a)

_L

10,0 m 5.0m \177t

B \342\200\242
\177B B

l\177zll\177rz2 I
(b)

FLmare 5.22. Example 5.15.

5.3.11 Stress distribution dt\177e to inter\177al loading

An approximation for the vertical stress component due to a vertical force in the interior of a
semi-infinite elastic soil was proposed by Geddes (i966 and 1969)\342\200\242
The approximation is
based on the Mindlin (i936) solution and follows a Boussinesq based approach. One impor-
tallt application is i\177 the evaluation of the vertical stresses beneath a pile due either to the sur-
face reaction in the form of an applied shear stress or the end reaction in the form of a single
concentrated load. Figure 5.23 shows the arrangement adopted for the solution. The vertical
stress component at a point A with a horizontal distance r from the load Q and a depth z > L is:

a: :\177-2
lq
(5.50)

where L is the depth of vertical load Q and

tq
is given by Equation 5.51. The coordinates of
point A are expressed in the dimensionless parameters of m = z/L. and n = r/L.)))
 Stress Distribution and Settlement in Soils 179

Fibre 5.23. Concentrated load applied in the interior of a semi-infintte elastic soil,

1 (1
-
2p,)(m
- 1) (1- 2bt)(m -
1) 3(m - 1)3
{ \"{\"
Iq - 2 3/\"\177- 2 3/2 - 1)215/2
871:(1 pt) [n + (m-t)2] [1\177 + (,,n +1)2 ] + (m
[,,\177\"\177

3(3 -
2 - 3(m + l)(5m - 3
4p.)m(m + 1) 1) 30m(m + 1)
-\177 + } (5.51)
2 5/2 2 7/2
[n +(m+l)-2] [n +(m +i)2]
where p, is Poisson's ratio of the elastic soil, If the applied load Q is transferred to the soil

through the surface area of the pile and the end reaction is ignored, the vertical stress compo-
nent is calculated from Equations 5,52 or 5.53. Equation 5,52 represents a uniform shear force
of magnitude Q / L developed along the length of the pile. Equation 5.53 corresponds to a lin-
ear shear forcestarting from zero at the top of the pile and increasing to a maximum value of
2Q / L at the bottom of the pile.

1/2 2 312 2
+ n2
(m 2+tl2) [n +(m-l)2] (m )3/2
,,, 2 + n2
4m(I +\177)(m + 1)(-- + _1)2 -(4m )
n 11
2
+ (m + 1)
2
[n ]3/2

6,n
2 2
(\177)6m[mn (,n+ i)\177 ]
12

2 5/2 2 5/2 (5.52)

(m +,,12) [,,i, +(m+l)2]

2(2 - la)(4m + 1) - 2{1- 21a)(m)

2
(m + 1)
t
- 2(2-1\177)
tq
4;\302\242(1_\177t){[n2
1/2 2
+(m-1)2] [ii +(m+l)2])))
180 Soil Mechanics: Basic Conceptsand Engineering Applications

- m3 - \177t)m
2tx) -8(2
+ (rn -
2(1 2
n2 mn l)3
1/2 3/2
(m 2+n2) [n 2+(m-1)2]
2
2m - 2 3
4m 3 (m + 1) + (m + 1)
- lSn 3
4p, n rn + 2(5 + 2tx)(rn)
n

In
2
+ (m + 1)2]3/2
2 3 3 2
2(7- 2\177t)mn -6m + 2(5 4-2\177.)(m)2m 6mn2(n -m2) + 12(m)2(m+ 1)51
2 2 2
(m + n )3/2 [n + (m + 1)

2 m 5 + 6ran2 2 _
m 2)
12(m) (,
+ 2 5/2
(rn +n2)

(m -
2+ 2 /2
+ 2(2 in(In
1)2 ]1 / 2 + m - 1 In 2
+ (m + 1) ]1 + m + l
(5
\177) }
2 1/2
+m
2 1/2
(m +t\1772) (m \177n2) +\177

Equations 5.51, 5.52 and 5.53 \177e presented in Table 5.2 for convenient values of
(Bowles,1996).
Example 5.16

Refe\177ing to Figure 5.24, compute the ve\177ical stress components at a depth of 2 m under each
pile for each of the following cases: (a) \177e load is ca\177ied by end being only, (b) one half of
the load is c\177ried by end bering and the other half by s\177n friction. = 0.3.
\177

Solution:
(a) z = 2.0+ 8.0= 10.0 m, m = z/L = 10.0/8.0= 1.25for all piles.
The ve\177ical stress component under pile 1 is \177uation calculated from
5,51 by superimposing
the effects of three loads according to: \177_ = + +
lq2 fq3) where lq2
and lq3 are the
(Q/L2)(lqt
co\177esponding effects of pile 2 and pile 3 at the point of interest under pile 1,
For m = 1.25, n = 0.0, from Table 5.2(a)and inte\177olating between 1.2 and 1.3:
=
lql (4.9099+2.2222)/2 3.5660.
=
For n = 4.0 / 8.0 = 0.5, lq2 = 0.1495, for n = 8.0 / 8.0 = 1.0. lq3
= 0.0506.

I\177
4.0m \177 4.0m

\177 2 3

Figure 5.24. Example 5.16.)))

 Stress Distribution and Settlement in Soils 181

Table 5.2(a). Vertical stress influence factors due to a vertical load applied in the interior of a soil.
p=0.2
n-\177 0,00 0.1 0,20 0.30 0.40 0.50 0.75 1.00 1.50 2.00
m,L
l.O ,= 0.0960 0.0936 0,08970,0846 0.0785 0.0614 0.0448 0.0208 0.0089
1.1 17.9689 1.7753 0.6182 0.2238 0.1332 0,0999 0,0659 0.0467 0.0222 0,0099
1.2 4,5510 2.7458 1.0005 0.3987 0,20560.1325 0.0724 0,0490 0,0236 0.0110
1.3 2.0609 1.6287 0,9233 0.4798 0.2672 0.1681 0.08110.0520 0.0249 0.0119
1.4 1.1858 1.0382 0.7330 0.4652 0.2926 0.1930 0.0905 0,0555 0.0263 0.0129
1.5 0.7782 0,7153 0.5682 0.4114 0.2875 0.2025 0.0985 0.0592 0.0277 0.0138
1\1776 0.5548 0.5238 0,4457 0.3518 0.2664 0.1997 0.1038 0.0625 0.02900.0147
1.7 0.4188 0.40180.3569 0.2984 0.2399 0.1893 0.10610.06510.0303 0.0156
1.8 0.3294 0.3193 0,2918 0.2539 0.2133 0.1775 0. I057 0.0668 0.0315 0.0164
1.9 0.2673 0.2609 0.2431 0.2177 0.1890 0.1606 0.1033 0.0675 0.0325 0.0172
2.0 0.2222 0.2\17780 0,2060 0.1883 0.1676 0.1462 0,0995 0.0673 0.0334 0.0179
Ix= 0.3

n-\177, 0.00 0.1 0.20 0,30 0,40 0.50 0.75 1.00 1.50 2.00

1,0 \177 0.1013 0.0986 0,0944 0.0889 0.0824 0,0641 0.0463 0.0209 0.0087
1.I 19,3926 3.9054 0.5978 0.2123 0,1287 0.0986 0.06680.0475 0.0222 0.0097
1.2 4,9099 2.9275 1,0358 0.4001 0.2027 0,1303 0,0722 0.0493 0.0235 0,0106
1.3 2.2222 1,7467 0.9757 0,4970 0.2717 0,1687 0.0808 0,0519 0.0247 0.0116
1.4 1.2777 1.1152 0.7905 0.4891 0.3032 0,1974 0.0908 0.0555 0.02600.0125
1,5 0.8377 0.7686 0.6070 0,4356 0.3012 0.2098 0.0999 0.0594 0.0274 0.0134
1.6 0.5968 0.5626 0.4768 0,3738 0.2809 0,2086 0.1063 0.0631 0.02880,0143
1,7 0.4500 0.43120.38190,3177 0.2538 0.1988 0.1094 0,0661 0.0302 0,0152
1.8 0.3536 0.3424 0.3122 0.2706 0.2262 0. t849 0.1096 0.06820.0315 0.0161
1.9 0.2866 0.2795 0.2600 0,23210.2006 0,1697 0,1076 0.0693 0.0326 0.0169
2.0 0.2380 0.2333 0.2201 &2007 0.17800.15470.1039 0.0694 0.03360.0177
\177= 0.4

n --\177 0.00 0.1 0.20 0,30 0.40 0.50 0,75 1,00 1.50 2.00
m

1,0 oo 0,1083 0.1054 0.t008 0.09470.0876 0.0676 &0483 0.0212 0.0083

1.1 21,2910 4,0788 0.5699 0.1970 0,12280.0970 0.0680 0,0486 0.0223 0.0093
1.2 5.3884 3,1699 1.0829 0.4020 0.1989&1274 0.0720 0.0496 0.0233 0,0102
1.3 2.4373 1.9040 1.0455 0.5200 0.27760.16950.0804 0.0519 0.0244 0.0111
1.4 1.4002 1.2\17779 0.8438 0.5208 0.3173 0.2032 0.0913 0.0554 0.0256 0.0120
1.5 0.9\17772 0.8395 0\1776587 0.4678 0.3194 0.2196 0.1017 0.0596 0.0270 0.0129
1.6 0.6527 0.6143 0.51810.4033 0,3001 0.2205 0.1095 0,0638 0.0284 0.0138
1.7 0.4915 0.4705 0.4152 0,34350.2724 0.2116 0,11380.0675 0.0300 0.0147
1.8 0.3858 0.3732 0.3393 0.2929 0,2433 0.1976 0.1148 0.0701 0.0314 0.0156
1.9 0.3123 0,3044 0.2825 0.25120.216t 0.18180.11330.07170,0328 0,0166
2\1770 0.2590 0.2537 0.2390 0.2173 0.1919 0.16590.1098 0.0722 0.0340 0.0174

c\177:l
=
0:3
= (tO00,O/8.0z)(3.5660 +0.1495 + 0.0506)= 58.8kPa,
Pile 2: tql = = 0.1495,
fq2
= 3.5660.
Iq3

\177z2
= (1000.0/8.02
)(0.1495 + 3.5660+ 0,1495) = 60.4kPa.)))
182 Soil Mechanics: Basic Concepts and Engineering Applications

Table 5.2(b).Vertical stress influence factors due to uniform shear force applied in the interior of the soil.

g=0.2
n --\177 0.00 0.02 0.04 0.06 0.08 O.1 O. 15 0.20 0.50 1.0 2.0
m,I,

1.0 \177,, 6.4703 3.2374 2.1592 1.6202 1.2962 0.8630 0.6445 0.2300 0.0690 0.0081
l.l 1,7781 1.7342 1.5944 1,4178 1.2418 1.0850 0.7953 0\1776138 0.2283 00730 0.0096
1.2 0.90t5 0.8789 0,8576 0,8269 0.7882 0.7446 0.6317 0.5307 0.22310.0759 0.0111
1.3 0.5968 0.5799 0.5725 0.56290.5500 0.5340 0.4867 0.4355 0.2138 00779 0.0125
1.4 0.4569 0.4288 0.4241 0.4201 0.4142 0.4068 0\1773838 0.3562 0.2010 0.0789 0.0139
1.5 0.3482 0.3359 0,3334 0,3313 0.3282 0.3242 0.3113 0.29520.18620.0790 0.0152
1.6 0.2922 0.2626 0.2716 0.2707 0.2689 0.2666 0.2589 0.2487 0.1708 0.0784 0.0165
1.7 0.25180.2304 0,2287 0.2274 0,2261 0.2247 0.2195 0.21270.15590.0770 0.0175
1.8 0.1772 0.\177953 0.1949 0.1942 0.1936 0.1925 0,1891 0.18440,1420 0.0750 0.0185
1.9 0.1648 0,1702 0,16980.16780.16820.16750.16500.1616 0.12950.0727 0.0193
2.0 0.1461 0.1482 0.1486 0.14800.14780.1473 0,1455 0.1429 0,1180 0.0700 0.0201
I.t
= 0.3

n ---> 0.00 0.02 0.04 0.06 0,08 0.10 0.15 0.20 0.50 1.00 2.00
m ,\177

1.0 \177, 6.8419 3.4044 2.2673 1.6983 1.3567 0.8998 0.6695 0.2346 0\1770686 0,0076
l.I 1.9219 1.8611 1.7072 1,51341.32111.1503 0.8368 0.6419 0.2335 0.0728 0.0091
1.2 0,9699 0.9403 0.9166 0.8825 0.8400 0.7922 0.6688 0.5588 0.2292 0.0760 0.0105
1.3 0.6430 0.6188 0.6099 0,5992 0.5850 0.5675 0.5157 0,4597 0.2207 0.07820.0120
1.4 0,4867 0.4558 0.4507 0.4461 0.4396 0.4316 0.4063 0.3761 0.2082 0.0796 0,0134
1.5 0.3766 0.3561 0.3533 0.3510 0.3476 0,3432 0,3291 0.3115 0,18340.0800 0.0148
1.6 0,3339 0.2895 0,2878 0,28630.2843 0.2817 0.2732 0.2621 0.1777 0.0796 0.0160
1.7 0.2664 0.2438 0.2414 0.2399 0.2384 0.2369 0\1772313 0.2239 0.1623 0.0784 0.0172
1.8 0.2025 0.2065 0.2054 0.2044 0,2038 0,2026 0,1989 0.19380.1479 0.0766 0.0182
1.9 0,1847 0,1794 0.1785 0.17770.17680.17600.1733 0.16960.1347 0.0744 0.0191
2,0 0.1634 0.1565 0.15610.1556 0.15510.15450.15250,1498 0\1771229 0.0718 0,0\17799

g=0.4

n --> 0.00 0.02 0.04 0.06 0.08 0.10 0.15 0.20 0,50 1,00 2\17700
m S

1.0 \177 7.2744 3.6270 2.4110 1.8026 1.4373 0.9488 0.7029 0.2407 0,0681 0.0069
1.1 2.0931 2.0296 1.8574 1.6409 1.4266 1.2372 0.8921 0.6794 0.2404 0.0725 0.0083
1.2 1.0486 1.0209 0.9947 0.95670.9091 0.8556 0.7181 0.5964 0.2373 0.0760 0,0098
1.3 0.6922 0.6694 0.6598 0,6476 0.6318 0,6122 0.5543 0.492\177 0.2298 0.0787 0.0113
1.4 0.5347 0,49220.4860 0.4807 0.4735 0.4645 0.4362 0.4026 0.21780.0805 0.0128
1.5 0.4020 0.3823 0.3798 0.37710.3734 0.3684 0.3527 0.3332 0.2029 0.0813 0.0142
1.6 0,3440 0.3096 0.3083 0,3068 0.3045 0.3017 0.2922 0.2800 0.1868 0.0812 0.0155
1.7 0.2943 0.2606 0.2580 0.2564 0.2549 0.2531 0.2469 0,2387 0.17080.0803 0.0167
1.8 0.2114 0.2207 0.2189 0.21810.2174 0.21610.2119 0.2063 0.1558 0,0787 0,0178
1.9 0.1782 0.19070.19040.18900,18810.[8730,1843 0.1802 0.1419 0.0766 0.0188
2.0 0.1741 0.16600.16580.16520.16480.16420.16200.15900.1294 0.0741 0.0196)))
 Stress Distribution and Settlement in Soils 183

Table 5.2(c). Vertical stress influence/'actors due to a linear shear force applied in the interior of a soil.
\177t= 0\1772

n -\177 0.00 0.02 0.04 0.06 0.08 0.\1770 0.15 0.20 0.50 1.0 2.0
m $

1.0 ,,\177 11.5315 5.3127 3.3023 2.3263 1.7582 1.03720.7033 0.1963 0.0618 0.0082
1.1 2.8427 2.7518 2.4908 2.1596 1.8329 1.5469 1.0359 0.7346 0.20740.0656 0.0096
1.2 1.2853 1.2541 1.2158 1.1620 1.0930 1.0162 0.8211 0,0110
0.6529 0.2141 0.0689
1.3 0,7673 0.7753 0.7585 0.7420 0.7195 0.6928 0.61420.53120.21390.07170.0123
1.4 0.5937 0,5450 0.5343 0.5267 0.5181 0,5063 0.4693 0,4261 0.2068 0.0737 0.0t36
1.5 0.4485 0.4051 0.4059 0.4006 0,3960 0.390t 0.3704 0.3460 0,19470.0750 0.0148
1.6 0.3635 0.3201 0.2326 0.31830.3154 0,3123 0.3008 0,2861 0.1803 0.0754 0.0160
1,7 0.3204 0.2583 0.2635 0.2618 0.2595 0.2574 0,2503 0.2408 0.1652 0.0750 0,0170
1.8 0.25330.2222 0.2239 0.2206 0.2181 0,2166 0.2122 0.20590.1506 0.0739 0.0180
1.9 0.2382 0.1761 0,1855 0.18800.18780.18530.18270.1782 0.13710.0722 0.0188
2.0 0.1776 0.1643 0.1648 0.1630 0.16310,16140.1591 0.1561 0.1248 0,0700 0.0196
\177.-- 0.3

n --\177 0.00 0.02 0.04 0.06 0.08 O.10 O. 15 0.20 0,50 1.0 2.0
m $

1.0 ,\177, 12.1310 5.5765 3.4591 2.4320 1.8346 1.07740.7276 0.1997 0.0616 0.0777
1.1 3.0612 2.9620 2.6751 2.3119 1.9547 1.6433 1.0908 0.7680 0.21150.0654 0.0090
1.2 1.3821 1.3465 1.3052 1.2465 1.17061.0864 0.8730 0.6899 0.2198 0.0689 0,0104
1.3 0.8262 0.8035 0.8130 0,7949 0.7705 0.7411 0.65480.5639 0.2212 0.0720 0.0117
1.4 0.6194 0.5827 0.5722 0.5630 0,5540 0.5410 0.5005 0.4530 0.21500.0744 0,0130
1.5 0.5189 0.4337 0.4332 0.42810.4227 0,4163 0.3946 0,3679 0.2033 0,0760 0.0143
1.6 0.3841 0.3415 0.3449 0.3395 0.3361 0.3327 0.3202 0.3039 0.1887 0.0768 0.0155
!.7 0.3332 0.2764 0.2810 0.2782 0.2764 0.2739 0.2660 0.2556 0.1732 0.07670.0166
1.8 0.2837 0.2268 0.2381 0.2347 0.2319 0.2300 0.22530.21830.15800.0758 0.0176
1.9 0.2654 0.1873 0,1963 0.19910.19880.1965 0,1937 0.1887 0,1439 0.0742 0,0186
2,0 0,18720.1730 0,1744 0.1732 0.1725 0.1714 0.1684 0.1651 0.1310 0.0721 0.0194
g=0.4
n --\177 0.00 0.02 0,04 0.06 0.08 0.10 0.15 0.20 0.50 1.0 2.0

1.0 ,\177 12.9304 5.9282 3.6683 2.5729 1.9365 1.13110.7600 0.2042 0.0614 0.0069
1.1 3.3525 3.2423 2.9209 2.5144 2.1171 1.7719 1.1641 0.8125 0.21700.0652 0.0083
1,2 1.5030 1.4712 1.4255 1.3588 1.2742 1.1800 0.9422 0.7394 0.2274 0.0689 0.0096
1.3 0.8965 0.9066 0.8862 0.8649 0,8383 0.8056 0.70890.6076 0.2308 0.0723 0.0109
1.4 0.6753 0.6350 0.6222 0.6120 0.6018 0.5874 0.5419 0.4890 0.22600.0752 0.0123
1.5 0.5629 0.4718 0.4712 0.4641 0.4584 0.4511 0.4270 0.3971 0.2147 0.0773 0.0136
!,6 0.4198 0.370I 0.3730 0.3672 0.3642 0.3600 0.3461 0.3278 0.1999 0.0786 0.0149
1.7 0.3752 0.2840 0.3039 0.3011 0.2984 0.2956 0.2870 0.2754 0.18380.0788 0.0161
1.8 0.3158 0.2496 0.2575 0.25300.2497 0.2479 0.2427 0.2349 0. t680 0.0782 0.0172
1.9 0.2851 0.2022 0.2122 0.2\17755 0.2142 0.2113 0.2083 0,2028 0.1530 0.07690.0182
2.0 0,2012 0.1929 0.1878 0.1854 0,1850 0.1837 0.18070.17710.13930.0749 0.0191)))
i84 Soil Mechanics: Basic Conceptsand Engineering Applications

5.4 ELASTIC SETTLEMENT OF FOOTINGS

5.4.1 Distribution of contact p ressureand settlement

Elastic settlement of the soil under a loaded area may be calculatedby integrating Equation
5.6 (vertical strain) along a vertical plane. The boundary conditions for the integration depend
on the rigidity of the footing that appIies the surface loading. In a rigid footing the setdement
under the loaded area is uniform. While the vertical stress component is independent of the
elastic properties of the soil, settlement, however, dependson the magnitudes
the eIastic of Es
and \177.Available analytical solutions have to be applied with care as the Modulus of Elasticity
may change with depth or with confining pressure. Analytical formulae and discussions for the
elastic treatment of heterogeneous soils are suggested by Poulos & Davis (1974), and Gibson
(i974). In general, a linear increase in the Modulus of Elasticity with depth is recommended.
In most clay soils it is reasonable to assume a constant Modulus of Elasticity, while in sands
the elastic formulation has Iittle use as the Modulus of Elasticity changes along the loaded area
as well as with depth. Figure 5.25 shows the distribution of contact pressure and settlement
under both rigid and flexible uniformly toaded footings constructed on clay and sand layers.
The assumption of ideal elastic properties for the soil implies that settlement occurs also out-
side of the loaded area and approaches zero in Boussinesq type behaviour. Most of the avail-
able formulations define the settlements at the corner of a rectangular loaded area as well as at
the centre and edges of a circular loadedarea. In the case of a rectangular loaded area, super-
position can be used to estimate the settlements within or outside the loaded area. Bowles
(1996)suggested that elastic solutions are reliable and the differences between measured and
calculated settlements are mostly due to the errors involved in the determination of the elastic
properties of the soil. An alternative approach is to divide the soil beneath the loaded area into
horizontal layers of finite thickness and estimate the vertical deformation of each layer. This
can be doneby estimating an average vertical stress component for each layer and calculating
its average vertical strain and deformation. This method has been found to yield reliable solu-
riot, s when compared with the traditionaI elastic numerical analysis (Powrie, 1997).

I I

Figure 5,25. Distribution of contact pressure and settlement: (a) contact pressure under a rigid footing,
(b) settlement under a flexible footing.)))
 Stress Distribution and Settlement in Soils 185

5.4.2 Elastic settlement under flexfble and rigfd footings

Based on the theory of elasticity (Timoshenko & Goodier, 1982),the elastic settlement Se of a
flexible footing, either rectangular of dimensions L \303\227
B (L > B) or circular of diameter B, is
given by:

Se =qB1-\1772 Is (5.54)
Es
where Is is an influence factor depending on the shape and L / B ratio. In this derivation, the
Modulus of Elasticity of the soil is assumed to be constant with depth and the soil is infinitely
deep. The influence factor for a flexible rectangular footing is based on the original analytical
solution developed by Steinbrenner (1934) for a corner of a loaded rectangular area.

I! = +1)+ (5.55)
\177[-\177ln( L/B

The centre of a flexible rectangular footing is the conunon corner for 4 rectangles of
L' = LI 2 and B' = B 12. Noting that L'I B\" = LI B, the settlement at the centre is:

Sc = 4 \303\227
qB'[(l
-
2)
\177t / E s ]I t = qB[(l - B2 ) / ),
Es ](211 and therefore:

I\177(centre)
= 2 It (5.56)
Thus the elastic settlement at thecentre of a flexible rectangular footing is twice that of its
corner. It can be shown that the average influence factor for a flexiblerectangular footing is
approximately 0.848 that of the corresponding influence factor at the centre:
=
I\177 0.848 x 21\177 and therefore:

ls (average) = 1.696I\177 (5.57)

For a rigid rectangular footing the influence factor is approximately 0.926 of the average in-
fluence factor of the flexible footing.
Hence, Is (rigid) = 0\177926 \303\227
Is (average)
= 0.926 x 1.69611,or:
Is (rigid)
= 1.5701t = (\177/2)11 (5.58)

The variation of Is with the L / B ratio is shown in Figure 5.26. For a flexible circular footing
the influence factors at the centre and edge of the footing are 1.0 and 2 / \177= 0.64 respectively,
and the average influence factor is 0.848. For a rigid circular footing the influence factor is
/
\177z 4 = 0.79 which is 0.926 of the average value for the flexible footing. Indeed, the factor
0.926 is the ratio of the settlement under a rigid circular footing with non-uniform contact
pressure distribution the
corresponding
to settlement to
the the uniform contact pressure under
same vertical load. This a rigid footing, regardless of the distribution
means that for of the
contact pressure, the maximum difference in the settlement is approximately 7.4%. The influ-
ence factor of 0.79 for a rigid circular footing is calculated on the basis of a non-uniform con-
tact pressure distribution. Note that the intluence factors for circular loading are basedon ana-
lytical solutions. It is evident that the ratio between rigid and elastic influence factors in a)))
186 Soil Mechanics: Basic Concepts and Engineering Applications

Cer tre
2.5

1.5
Cor\177er

1
I

0.5
I Avlrage (circ
0 Rigid (ci \"tie) ge ircle]
(\302\242

1 2 3 4 5 6 7 8 9 10
L/B
Figure 5.26, Steinbrenner's influence factors for circular and rectangular footings.

circular footing is equally applicable to a rectangular footing. Based on an analysis carried out
by Jumikis(1969), the influence factors within and outside a flexible circular footing are pre-
sented in Table 5,3. The parameter a representsthe distance of the point of interest from the
centre of the circular footing whilst Rc is the radius of the circle.

Table 5,3. Influence factors for a flexible circular footing.

a / Rc ts a / Rc Is a / Rc Is
0.00 1.000 1.75 0.307 3.50 O. 146
0.25 0.984 2.00 0.258 3.75 0.134
0.50 0.934 2.25 0,234 4.00 0.125
0.75 0.840 2.50 0.204 4.25 0.118
1.00 0.637 2.75 0.191 4.50 0.112
1.25 0.444 3.00 0.167 4.75 0.105
1.50 0.360 3.25 0.157 5.00 0.099

Example 5.17

A 4 m by
8 m flexible rectangular footing exerts a uniform contact pressure of 200 kPa on the
ground surface. Calculate the elastic settlement at its centre, cornersa\177d the centre points of its
long and short sides. Es = 20000 kPa, g
= 0.35.

Solution:
L / B = 2.0,I, = 1.53 for the centre point and fs = 0.76 for the conters. From Equation 5.54:
Se (comer) = 200x4.0x[1-
2 = 0.0267
(0.35) ]x 0,76/20000 m.
Se (centre) = 200x4.0x[1 - (0.35) 2 ]xl.53/20000 = 0.0537m. To find the settlement at the
centre point of the longer side the original footing is sub-divided into two rectangles (in this)))
 Stress Distribution and Settlement in Soils 187

example two squares) sharing the centre point at their COln\177ers. The total settlement at this
point is equal to twice the settlement of the comer of one rectangle:
L/B = 4.0/4.0 = 1.0. I\177= 0.56,
Se (centre of long side) = 2\303\227200\303\2274.0x[1 -(0.35)2]\303\2270.56/20000= 0.0393 m.
For the centre point of the short side L //3' = 8.0 / 2.0 = 4, Is = 0.98,
Se (centre of short side) = 2 \303\227 -
200 x 2.0 \303\227 = 0.0344 m.
[1 (0.35)2 ] \303\227
0.98/20000

Example 5.18
Find the elastic settlement at point A of the flexible footing shown in Figure 5.14.
Es = 15000kPa, Ix
= 0.3.

Solution:
The right half of the footing is divided into 5 rectangles all sharing point A as a common cor-
ner. The results of calculations are tabulated below.

Rectangle L (m) B (m) L 1 t? 1\177 Se {m)

Aabc 4.0 2.0 2.0 0.76 0.0277

Adef 3.0 3.0 1.0 0.56 0.0306

Adgc 3,0 2.0 1.5 0.68 -0.0247
Ahif 3.0 1.0 3.0 0.89 -0.0162
Ahjc 2.0 1.0 2.0 0.76 0.0138
+ 0.0306
S,,=2(0.0277 - 0.0247 - 0.0162+ 0.0138)= 0.0624 m

5.4.3 Settlement of a soil layer of finite thickness

If an incompressible layer (for which Es --\177\177,) is located at depth H from the grouad surface,
the surface settlement is reduced.In this case, the I\177values for a flexible rectangular footing
are given by:
-
\177 2g
I s (corner)= I1 =
I2 + I3, I.\302\242
(centre) = 2I
1, I s (average) = 1.69611 (5.59)
where 12 and I3 are Stelnbrenner's influence factors given by:

I\177
\"

[. L/B(\177 + if(L/B)
\177
+ (H / B) \177+ \177)
=\177[\177

_ (H/B) L/@
13)))
tan_\177[ .] (5.61)
188 Soil Mechanics: Basic Concepts and Engineering Applications

Figure 5.27. Steinbrenner's influence factors for the elastic settlement of a layer of finite thickness.

Equation 5.59 (Is (centre)) may be used for the centre of a flexible circular footing by convert-
ing the circle into an equivalent square. The influence factor at the edge of the circle is 2 / x
times I.\177(centre). For rigid rectangular and circular footings the influence factor is 0.926 of the
average, or 0.79 of the central values. Figure 5.27 shows the variations of 12 and I3 with H / B
for selected values of L / B (circled numbers).

Example 5.19

Re-work Example 5.17 with a rigid layer located at i2 m depth from the ground surface and
compare the resuIts.

Solution:

H / B = i2.0 / 4.0= L 3.0, / B = 2.0,

substituting these values into Equations 5.60 and 5.61:

12 = 0.402, 13 = 0.084,thus from Equations 5.59:

I\177(comer)
= 0.402 + (1 - - 0.35)
2x 0.35)/(i x 0.084 = 0.4408.

Is (centre) = 2 x 0.4408 = 0.8816.

Se (corner) = 200x 4.0x|1- (0.35) 2

]x 0.4408/20000 = 0.0155 m.
Se (centre) = 200x4.0x[1 - (0.35) 2
]x0.8816/20000 = 0.0309 m.
For the centre point of the long side: H / B = 3.0,L/ B = 12
= 0.363, 1.0, 13 = 0.048,
I s = 0.363+(1- 0.35)x
-
2x0.35)/(1 0.048 = 0.3851.
2 x 0.3851/ 20000= 0.0270m.
200x 4.0x [1
Se (centre of long side) = 2x ] -
(0.35)
For the centre point of the short
-
side: H / B 6.0,L / B = 4.0, lz = 0.609,I3=0.087,
Is = 0.609+ (1- 2x 0.35)/(1- 0.35) x 0.087 = 0.6491.
Se (centre of short side) = 2x 200x2.0x[1- (0.35)2 ] x0.6491/20000
= 0.0228 m.

The settlements significantly reduced by the presence of the incompressible layer.)))

 Stress Distribution and Settlement in Soils 189

5.4.4 Reduction of settlement due to footing embedment depth

Fox (1948) investigated the effect of embedment the of a footing (D) on the average elastic
settlement of a uniformly Ioaded rectangular area and introduced the concept of a depth factor.

Here, the average settlement for the footing at ground surface level is muItiplied by the depth
factor 1p which is always Iess than unity. PIots of the depth factor are given in Figure 5.28.
From Figure 5.28(b) it can be seen that the depth factor decreases with a reduction in Poisson's
ratio. Table 5.4 shows the variation of IF for \177= 0.5 for various values of L1B and D 1 B.

Table 5.4. Depth factors IF for p.

= 0.5,

D/B L/B=I.O L/B=2.0 L/B=3.0 L/B=4.0 L/B=5.0 L/B=tO.O

0.00 1.0000 1.0000 1,0000 1.0000 1.0000 1.0000
0.25 0.9424 0.9650 0,9723 0.9760 0.9783 0.9883
0.50 0.850I 0.8993 0.9184 0.9284 0.9348 0.9489
0.75 0.7760 0.8372 0.8651 0.8808 0.8909 0.9139
1,00 0,7234 0.7864 0.8196 0.8394 0.8525 0,8829
1,50 0.6587 0.7150 0.7510 0.7750 0.7919 0.8334
2 0.6220 0,6698 0.?039 0.7288 0.7473 0.7957
3 0.5828 0.6180 0.6459 0.6684 0.6868 0.7412
4 0.5625 0,5899 0.6126 0.63\1779 0.6485 0.7030
5 0.5502 0.5724 0.5913 0.6079 0.6226 0.6746
6 0.5419 0.5606 0.5767 0.59tl 0.6040 0.6526
8 0.5315 0.5456 0.5580 0.5692 0.5796 0.6210
l0 0.5252 0.5366 0.5466 0.5557 0.5643 0.5998
50 0.5050 0.5073 0.5094 0.5113 0.5131 0.5212
100 0.5025 0.5037 0.5047 0.5056 0.5065 0.5106

0.95
0.9 ----\177
0.85 ! !\" --
0.8 , :

o. 7 5 \\\\%,
0.7 __._.\177.p.=0.
0.65 , ,

0.6 : :

0.55 , ,

0.5 , ,

0.45

0.1 1 10 10)))

D/B D/B
(a) (b)

Figure 5,28, Depth factor: (a) saturatcd clay, p,

= 0,5, (b) effect of Poisson's ratio.
190 Soil Mechanics: Basic Concepts and Engit\177eering Applications

0.95

0.9
0.85
0.8
0 5 10 15 2O
D/B
2

1.5

ptl

0.5
Circle

0.1 1 10 100 1000

H/B

Figure 5.29, Improved influence factors \177 and \177hfor saturated clays (Christian & Carrier. 1978).

For a circular footing an equivalent square area may be used. Fox's depth factor may also be
applied to consolidation settlement as recommended by Tomilinson, (1995),

5.4.5 Alternative influence factor.r for saturated clays

For shallow and deep f]exible footings in saturated clay with p. = 0.5, the chart proposed by
Janbu et al. (1956)finds significant use in design. The settlement is caIculated from:

S e = go\177ttqB/E s (5.62)

where I.t0 and IUt| are influence factors for depth and layer thickness respectively, hnprove-
merits to the \177t0and I.tL influence factors were proposedby
(Figure 5.29) Christian & Carrier
(1978) based on the work carried out by Burland (1970) and Giroud (1972).

ExampIe 5.20
A flexible 2 m x 2 m footing is located at a depth of 4 m and exerts a uniform contact pressure
of 400 kPa. If an incompressiblestratum is at 10 m below the ground surface, calculate the av-
erage elastic settlement of the footing. Es = 20000kPa, p.
= 0.5.)))
 Stress Distribution and Settlement in Soils 191

Solution:
H / B = 10.0/ 2.0= 5.0and L / B = 2.0 / 2.0 = 1.0,I2 = 0.437, I3 = 0.031,
=0.4370.
I s (comer)= 0.437+ (1 - 2 x 0.5)/(1 - 0.5)x 0.031
1 (average)
s
= 1.696\303\227 I
s (corner)
-- 1.696 x 0.4370 = 0.741i.
For D / B= 4.0 1 2.0 = 2.0,the depth 5.4 is: 0.622,
factor from Table
Se(average) = 400x 2.0x[1- lx0.622/20000 = 0.0138m.
(0.5)2]x0.741
Using Janhu's improved chart (Figure 5.29)for H / B = 3.0 and L / B = 1.0: go= 0.9,P-L \177
0.58,
From Equation 5.62: Se = 0,9x0.58x 400x 2.0/20000 = 0.0208 m.

5.4.6 Estbnation of settlements in sands

Due to the variation compressibility characteristics of sands,Equation

in the 5.54 is not appli-
cable to the estimation elastic settlement. This non-uniformity
of their exists under identica\177
footings in the same layer of sand as well as in laboratory samples. This non-uniformity be-
comesreasonably acceptable for normally consolidated sands. However, as discussed in Chap-

ter 4, the stress level in the sand should be more than 700 kPa in order to approach to the rele-
vant NCL (Atkinson & Bransby, 1978), and recompression in the moderate stresses is a
horizontal line indicating no volume change. The following methods used to estimate settle-
merits in sands are based on in-situ tests and the theory of elasticity.

Esthnatian of settlement using plate load tests.Terzaghi & Peck (1948) reported an extrapola-
tion equation that related the settlement of a loaded circular or square plate of 0.305m (S0.305)
to the settlement of a tboting (Se) of width B with the same contact pressure.

Se = S0.305 2
[2B/(0.305 + B)} (5.63)
The unreliability of this
extrapolation equation is discussed by Bjerrum & Eggestad (1963),
Sutherland (1975),
Clayton et al. (1995) and Terzaghi et al. (1996). I\177owever\177 recent findings
in cemented residual soils show that extrapolation can be carried out to relate the settlement of
a loadedplate to the settlement of an actual footing. Consoli et al. (1998)ha','eshown that, by
expressing the load test results in terms of dimensionless variables (e.g. normalised applied
stress versus ratio of settlement to diameter / width), clear behavioural patterns emerge.

Estimation of settlements using the Burland method. Based on the results of

and Burbidge
Standard Penetration Test (SPT; Chapter 10) as well as observations and measurements of
numerous field cases, Burland &/3urb/dge (1985) suggesteda semi-empirical solution for the
settlement of a normally consolidated sand. The surface settlement (Se),which is equal to the

integration of vertical strain \177z, is estimated using the average value of the coefficient of com-
pressibility m\177 within the strain influence depth of Z; under a surface loading of q:

Se = f\177t gzdz
= Z
lmvq (5.64)

The parameter my is defined by: mv =

8.. (average) / or: (average).)))
192 SoilMechanics:BasicConcepts and Engineering Applications

Empirical expressions for Z! (in m) and my (in MPa-l) are:

ZI = B 0'75 (5.65)

mv = 1.7/\1771.4
(5.66)
where B is the width of the footing in m and N is the average SPTnumber within the influence

depth Z1. The value of N is corrected for saturated sands and gravelly sands according:
N\" = 15 + 0.5(N - 15) saturated sand (5.67)

N' = 1.25N gravelly sand (5.68)

For a normally consolidated sand the general equation for its settlement is:
1.7
Se =B\302\260-75 (5.69)
\177-q

where Se is the settlement and q is the

in mm contact pressure in kPa. Note that Equation 5.69
does not imply a linear relationship between the settlement and the width of the footing, as any
change in B will alter ZI which in turn will change the average SPT number N. For overcon-
solidated soils, and for q greater than the preconsolidation pressureP'c,the settlement is the
sum of the recompression and normaIly consolidated stages, and a different value for my at the
recompression side has to be evaluated.Assuming the compressibility index at the recompres-
sion stage is 1/ 3 of the normally consolidated stage, then:

= B \302\26075 - for q > P'c (5.70)

Se 2pc / 3)
-\177-(q

If the applied contact pressure is less than P'c this becomes:

1.7 ,
Se = BO'75 / 3) for q <p c (5.71)
\"\1777-\177-(q

Equations 5.69 and 5.70 are also applicableto footings located at some depth below the
ground surface if the preconstruction verticaI stress is substituted for P'o For rectangular foot-

ings with L / B> 1.0,the settlement is increased by a shape factor given by:

=
I s [(1.25L/B)/(L/B+0.25)]2
(5.72)

The maximum value of this factor occurs in a strip footing when LIB--\177 and is equal to
1.56. If the incompressible layer is located within the influence depth Zt, the calculated settle-
ment is multiplied by a correction factor defined by:

If =H(2-H/Z t )/Z t (5.73)

where H is the depth of the incompressible layer from the footing level.

Example 5.21
A footing on sand has a pIan dimensions of 2 m x 6m and exerts a loading of q = 300 kPa.)))
 Stress Distribution and Settlement in Soils 193

The values of N from SPT tests are tabulated below. If the water table is I m belowthe ground
surface, calculate tl\177e settlement of the footing for the following cases: (a) the base of the foot-
ing is on the ground surface, (b) the base is 1 m below the ground surface. For both cases find
the uniform contact pressure that causes 25 mm average settlement.

Depth (m) 0,0 0.5 |.0 1.5 2.0 2.5 3.0

N 11 12 14 19 19 23 27
Solution:
The correctedN values (for N > 15 and in the saturated zone) are tabulated below:

Depth (m) 0.0 0.5 1.0 1.5 2.0 2.5 3.0

N It |2 14 17 17 19 21

From Equation 5,65 Zt

= B075 = 2.0\302\260'75
= 1.68 m; for case (a) the average value of \177\"=14,

thus from Equations 5.69 and = 1.68\303\2271.7 \303\227300.0/1414 = 21.3 mm,

5.72:S(2.0\303\2272,0 )
2 =
Is =[1.25\303\2273.0/(3.0+0.25)] 1.33, S(2.o\303\2276.0)
= I = 21.3\303\2271.33
\303\227
S(2.o\303\227:\177.o) s
= 28.3 ram.
\177'4=
For case (b) the average value of \177 =17, and = 1,68\303\2271.7\303\227300.0/17
S\177:z.0\303\2272.o)
16.2 ram,

S(2.0\303\2276.0)
=
S(2.0x2.0)
\303\227 = 16.2x
Is
1.33 = 21.5mm.
For 25 mm settlement: =
1.33xl.68xl.7xq/141'4 = 25 ram, q = 264.8kPa for case
S(2.0\303\2276.o)

(a) and =
1.33xl.68xl.7xq/1714
= 25 ram, q = 347.5kPa for case (b).
S(2 o\303\2276.o)

Estimation of settlements using Schmertmann's method. An alternative method of estimating

settlement is based on a numerical approximation of Equation 5.64 in which the soil beneath
the loaded area is divided into a finite number of layers:
i=n
= (5.74)
Se \177 (\177zz\177)i
i-I
where n is the number of the layers within an influence depth of Zt, (zt)i is the thickness of a
typical layer and (ez)i is the average vertical strain of the layer. Substituting (\177z)i from Hooke's
law (Equation 5.6) into Equation 5.74 we obtain:

+cry)
zt (5.75)
i=t\177_ E, i

Schmertmann (1970) and Schmertmann et al. (1978) proposedsolutions for sands in which
there was no need to estimate the stress components in the vicinity of the influence depth. The
Modulus of Elasticity must be determined for each layer by means of in-situ testing, preferably
using the Cone Penetration Test (CPT;Chapter 10). Introducing a strain influence factor Iz to
replace the stress term in Equation 5.75, tl\177e settlement increment is defined by:

lz =
[crz -g(crx +
c\177y
) ]/ q --\177AS
e
= I
q\303\227
z
/ Es (5.76))))
194 SoilMechanics:BasicConcepts and Engineering Applications

Equation 5.75 may therefore be written as:

Se= q (5.77)
+--\177sZl

The influence depth Z; is defined by:

Z! = 2B[I + Iog(L/B)] (5.78)

In a circular loaded area, the strain influence factor lz under its centre can be calculated
mathematically in terms of depth z or relative depth z / B (B being the diameter) by integrating
\177zover the influence depth assumed under the footing. The variation of Iz with depth has a
Boussinesq pattern and, after increasing to a maximum, decreasesto a negligible value in the

specified influence depth. Mayne & Poulos (1999)presented the strain influence factors for a
circular loadedarea for both rigid and flexible footings. Papadopoulos (1992)introduced a dif-
ferent model for the distribution of applied stress which takes into account the density of the
soil; strains can then be integrated to yield the settlement of the footing. However, depending
on the shape of the footing, the mathematical integration of e: is either difficult or, in some
cases, impossible. Strain influence factors, given by Schmertmann (1970) and Schmertmann et
al. (1978), that idealize the Boussinesq pattern by straight lines (e.g. Line ABC in Figure
5.30(a)) are based on extensive in-situ tests, and are shown in Figures 5,30, Equation 5.77 is
multipIied by depth and time factors defined by:

Ct = 1- t q) >_ 0.5 (5.79)

0.5(p\177

C2 = 1+0.2log(t/0.1) (5.80)

Strain influence factor Strain influence factor

0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
0 0

0.5 0.5 L
1 1 Axisym me\177tric,,,,,

1.5 1.5
2 2
2.5 2,5
3 Plane strain
3
3.5 3.5
4 4

4.5 4.5

Figure 5.30. (a) Strain influence factor (Schmenmann, 1970), (b) improved strain influence factor
(Schmertmann, et al., 1978).)))
 Stress Distribution and Settlement in Soils 195

where p'o is the effective overburden pressure at the footing level and t is the elapsed time in
years. Correction of the settlement for the effects of time may not be necessary (Holtz, 1991:
Terzaghi et al., 1996) as observed field results are probably due to consolidation of clay or silt
lenses within the sand. The term q represents the net contact pressure at the foundation level.
If the footing is at some depth below the ground surface, q has to be replacedby q -p'o. In the
improved strain influence factor diagram the maximum value of t: is given by:
t
:(max) =0\"5+O'I\177/(q-P'o)/P'(B/ZorB)
(5.81)
where P'(B/ 2 or B> is the effective overburden pressure at the depth of B / 2 or B from the foot-
ing level depending on L / B = i (axisymmetric)
the L / B ratio. For and L / B > i0 (plane
strain), the effective overburden pressures correspondsto B/2 and B respectively.

Example 5.22
A flexible rectangular footing of 2 m x 4 m applies a uniform load of 100 kPa to a horizontal

ground surface. The underlying stratum is divided into 6 layers as shown in the table below.
Estimate the elastic under the centre of the footing
settlement for the following cases: (a) the
soil has a constant Modulus of Elasticity of Es = 15000 kPa. (b) the soil is heterogeneous and
its Modulus of Elasticity changes linearly from 15000 kPa at the \177ound surface to 30000 kPa
at adepth of 10 m.

1 \"\177 3 4 5 6
Layer (downwards)
Thickness (m) 0.50 1.00 1.00 1.67 1.67 3.00

Solution:

The influence factor for the vertical stress or: at the centre of each layer is 4 times that for the
influence factor calculated for the corner of L' = L / 2 = 4.0/ 2 = 2.0m and B\" = B / 2 = 2.0/ 2
= 1.0m. The vertical strain a: is computed
average from Equation 5.6 by ignoring the lateral
normal For case (a) e: = c\177:l 15000.
stresses.
-
For case (b) E\177 Eo+ nz, where \302\243o
= 15000 kPa and n = (30000 - 15000) / I0.0 = 1500.thus:
=
E\177 15000 + 1500a and \177_ = or:/(15000 + 1500z). In both cases:
-,-,M = the settle,nent of each layer
=
e: \303\227
zt, where zt is the thickness of each layer.
The results of computations for both cases are tabulated below in which the symbols C and L

represent constant and linear E\177respectively.

\177(C) \177xs(L)
Layer Depth(m) m n 4xlq (kPa) 1 0-
x\177l (into) (into)

0.25 8000 4.000 0.99343 99.34 6.62 6.46 3.31 3.23

1.00 2.000 1.000 0.79976 79.98 5.33 4.85 5.33 4.85
2.00 1.000 0.500 0.48070 48.07 3.20 2.67 3.20 2.67
3.33 0.600 0.300 0.25174 25.17 1.68 1.26 2.80 2.10
5.00 0400 0.200 0.13119 13.\1772 0.87 0.58 1.45 0.97
7.34 0.272 0.136 0.06562 6.56 0.44 0.25 1.32 0.75
Tolal: 17.4 14.6)))
196 Soil Mechanics: Basic Concepts and Engineering Applications

Example 5.23

A square footing 3 m \303\227

3 m is located of 2 m and exerts
at a depth a uniform contact pressure
of 200 kPa (includes the weight of the the footing level) on an underlying
soil above deep de-
posit of sand.Calculate Se using Schmertmann's modified strain influence factor if the varia-
tion in Es with depth z given by: Es (kPa)
= 5000 + 3000 z (m). p = 1.7 Mg/m3.
Solution:

P'o
= 1.7 9,81
\303\227 2.0
\303\227 = 33.3 kPa, q - P'o= 200.0- 33.3 = 166.7 kPa.
From Equation 5.79: Cl = correction factor for depth = 1 - 0.5(33,3 =0.9,
/ 166.7)
Calculating the influence depth from the footing level (Equation 5.78):
Zt = 2\303\227
[1 +
3,0 \303\227 Iog(3,0/3,0)] = 6,0 m. From Equation 5.81:
= 0.5 + i] =
+ (3.0/2)\303\227i.7\303\2279.8 0,669.
Iz(max) 0.1-\177166.7/[33.3

For L / B = 1.0,the maximum strain influence factor occurs at the depth of 3,0 / 2 = 1.5m
from the footing level or 1.5+ 2 = 3,5 m from the ground surface. The equation of the line rep-
the strain influence factor from the foundation level to the of 3.5 m is:
resenting

(I - O. 1) t(z - 2.0) = (0.669

z
- 2,0), or I -
0,1)/(3.5
depth
= 0,1 + 0.3793(z- 2.0).
z
The strain influence 3.5m to z = 6.0 + 2.0= 8.0m
factor for the range z = is:

(I z
- - =
0.0)/(z 8,0) (0.669 8.0),
-
- 0,0)/(3,5
or 1: = 0.1487(8.0- z).
The two equations above are plotted in Figure 5.3 i. The influence depth is divided into 6 lay-
ers, each i m thick. Results of the calculations are tabulated below.

Layer
D
I,\" Es 1z x Zl / Es
Layer
D
1,\"
Es I: x zl/ Es
(rn) (kPa) (nmv\177Pat (rn) (kPa) (mm/kPa)
1 2.5 0.2896 12500 0,0232 4 5.5 0.3717 21500 0.0173
2 3.5 0.6690 15500 0.0432 5 6,5 0.2230 24500 0.0091
3 4.5 0.5204 18500 0,0281 6 7.5 0.0743 27500 0.0027
Total 0.1236

From Equation 5.77, Se =0.9 x 166.7x 0.123618.5 = mm,

5.4,7 Estimation of the elastic properties of soil

The Modulus of Elasticity Es can be estimated
the slope of the experimental from stress-strain
relationship obtained from an compression test. The tangent
appropriate triaxial modulus is the
slope of the stress-strain curve at the early stage of a test whilst the secant modulus is the slope
of a line from the origin to a point at 50% of the peak strength (Figure 5.32). Poisson's ratio
can be obtained using Equation 5.8, as at different stages of the triaxial compression test both
axial deformation and volume change are recorded.Samplesmust accurately represent the
field condition in terms of physical properties such as density and moisture content. Any dis-
turbance to the specimen will decrease its Modulus of Elasticity (Bowles, 1996), However, a
sample with moisture content drier than that in the field may give a higher value for Es. In
granular materials where the preparation of an undisturbed sample is not possible, care has to
be taken to duplicate the void ratio during the preparation of a disturbed sample, It is vital that)))
 Stress Distribution and Settlement in Soils 197

Strain influence factor

0 0.2 0.4 0.6 0,8

Figure 5,31. Example 5.23,

any disturbed or undisturbed sample be

preparedin accordance with the relevant code. In
granular materials the confining
pressure (63) should be chosen to match field conditions. An
increase in the confining pressure increasesthe Modulus of Elasticity of the specimen. The
drainage condition in the test must also simulate the field situation, especially in clay soils. It
is always good practice to determine E\177 in both the fully drained and undrained conditions.
Based on laboratory and field experiments, D'Appolonia et al. (1971) reported that the
Modulus of Elasticity in clay varies with plasticity index, sensitivity, and overconsolidation ra-
tio, and may be formulated as a linear function of undrained shear strength. Furthermore, the
Modulus of Elasticity of clay from different places around the world varies from the very
small value of 100 kPa up to about 13000 kPa. Table 5.5 shows the variation of the ratio of
undrained modulus to undrained shear strength in terms of overconsolidation r atio and plastic-

ity index (Duncan & Buchignani, 1976). In sands the magnitudes of the elastic parameters

ql
- \177r3
Initial tangent
modulus
// / Secant modulus

.... 50% --
((\1771 O'3)max

\177
\177--

Figure 5.32. Determination of Modulus of Elasticity.)))

 198 Soil Mec#anics: BasicConcepts and Engineering Applications

Table 5.5. Variation of Esu/'r. l in clay soils (Duncan & Buchignani, 1976).

OCR Pf < 30 30< Pf <50 Pt > 50

<3 600 300 125

3-5 400 200 75
>5 t50 75 50

depend upon size as well

the as the initial
particle void ratio (Hart, 1966 and 1977), The
Modulus reduces with decreasing of size of particles.
of Elasticity Lower values of initial void
ratio yield a high value of 45200kPa for Es in coarse and medium coarse sand (Table 5.6),

Table5.6.Modulus of Elasticity and Poisson's ratio in sandy soils (Haw, 1966).

Poisson's E\177(kPa)
Soil ratio: \177t 0.41 <e <0.5 0.51 < e < 0.6 0.61<e<0.7
Coarse and medium coarse sand O.15-0.2 45200 39300 32400
Free-grained sand 0.25 36600 27600 23500

Sandy silt 0.3-0.35 13800 11700 10000

Recent improvements in the correlation of SPTand CPT results to the soil strength properties,
the Modulus of Elasticity and compressibility index have created useful practical relationships.
Bowles (1996) suggested the f\177bllowing empirical retationships between SPT results and

E\177= 500 (N + 15) for normally consolidated sands (5.82)

for overconsotidated
sands (5.83)
Es =250 (N+ 15) for saturated sands (5,84)

Es
= 600 (N + 6) for gravelly sands, N _< 15 (5.85)

E\177
= 600 (N + 6) + 2000 for gravelly sands, N > 15 (5.86)

Es = 320 (N + 15) for clayey sands (5.87)

Es= 300 (N+ 15) for silts (5.88)

Using CPT results, Meyerhof & Fetlenius (1985) proposed the following relationship in which
the term qc represents the tip resistance that has the same units as Es.
=
Es Kqc (5.89)

where: K= 1.5 for silts and sands, K = 2.(? for compacted sand, K= 3 for dense sand, and K
=4 for sand and gravel. Schmertmann et al. (1978) suggestedthe use of the following for axi-
symmetric and plane strain loadings:

Es = 2.5qc for axisymmetric (5.90)

Es= 3.5qc for plane strain (5,91))))

 Stress Distribution and Settlement in Soils 199

5.5 SOIL-FOOTING INTERACTION MODELS

5.5.1 Tf\177e concept of rigid footing

Figure 5.33(a) shows a rigid footing subjected to parallel vertical forces and moments that are
applied about the axes in the plane of footing. This system is equivalent to a single resultant

force P (Figure 5.33(b)) that more convenient to

is perpendicular to the plane of footing. It is

express the contact

pressure system in terms
Gxyz where of
G representsthe
the coordinate
centroid of the footing, z is perpendicular to the footing and parallel to the contact pressure,
and xGy is the plane of the footing. To simplify the problem the resultant force P is moved
from its application point O to point G, which requires the addition of the two external mo-

ments, Ms and My, to maintain equilibrium:

Mx =
- P YO. = P xo (5,92)
My

where xo and YO are the O (assumed positive in Equation

coordinates of point 5.92) in the xGy
system, and theright-hand sign for moments
rule is adopted. The
conventioncontact pressure
under the rigid footing is assumed to be linear and independent of the type of soil:

q = Ax + By + C (5.93)
where A, B and C are coefficients depending on the geometry of the footing and applied
forces, and x, y are the coordinates of any point within the contact area. Thesecoefficients are

found by solving the three equilibrium equations of \177

F
z
=0. \177
M x =0 and \177.M y
=0:

lrvM \177+ IrM

+ P
\"\177 \"
\342\200\242 v
\" l\177.Mv
\"\" \" lyM \"r
A= , B= , C=\177 (5.94)

where S is the (plan) area of the footing,

contact fx and ty are the second moments of area of
the footing about the x and la\177,is the product second moment of area in the xGy
and y axes,
plane. If the application point of the resultant P coincides with the centriod of the footing, then
Mx = My= 0 and the soil pressure is uniform and given by q = P / S. If the footing is symmetri-
cal about the Gx or Gy axes, then l.\177.= 0 and:

P
x - Mr\"
My
q= tx
\"
S
(5.95)
ly

(b) (c))))

Fi\177ware 5.33. Rigid footing subjected to vertical forces and moments.

200 Soil Mecfvanics: Basic Concepts and Engineering Applications

A negative contact pressure implies tensilestressesunder the footing. The position of the zero
pressure line (where the contact pressure is zero) can be determined by setting Equations 5.93
or 5.95 to zero. If this line is located outsideof the soil-footing contact area then the contact
pressure under the footing is compressive. Figure 5,34(a) shows a plan of a rectangular rigid
footing in which bc is. the zero pressure line (ZPL), The triangle abc represents the tensile
stress appliedto the soil beneath the footing; the rest of the contact area is subjected to com-

pressive stress. The vertical force equilibrium equation may be written as:

P=Fc-F '
where Fc is the soil reaction under the compressed area and Ft is the tensile force correspond-
ing to area abc. The soiI cannot tensile stress and therefore Ft approaches
sustain zero. In order
to maintain equilibrium, the position of the zero pressure line has to be adjusted by reducing
Fc to satisfy P = Fc and moving the application point of the new Fc to point O. It may be
shown that a linear equation of the following form satisfies the equilibrium requirements.

q
= Mh I l (5.96)
where q is the contact in the compressive area under
pressure at any point the footing, M is the
moment of the resuhant about zero pressure Iine (R-ZPL), h is the distance of
the readjusted
the point from the new zero
pressureline and I is the second moment of area of the compres-
sive area about the readjusted zero pressure line. The position of the new zero pressure line is

usually found by iteration (Peck et oh, i974) until Equation 5,96 is satisfied. Note that a part
of the footing will be ineffective in resisting the applied loading. For rectangular footings, Irles

m)))

F\177.gure 5.34. Determination of zero pressure line.

 Stress Distribution and Settlement in Soils 201

& Ides (1994) suggestedan explicit for the position of the zero pressure
solution line. If the
application point of the resultant the x-axis, as shown
is on in Figure 5.34(b), the zero pressure
line will be parallel to the y-axis. The new position of the zero pressure line (point j in Figure
5.34(c)) is found by drawing the ]inejn to satisfy equilibrium in z direction:

(ll2)x jdxdnxB =P
The resultant P must pass through the centroid of trianglejdn:

jd =3(Ll2-xo)
Thus the maximum soil pressure dn is:

= 2P/[3B(L/2- x 0)] (5.97)

qrnax

For a rigid rectangular footing with the symmetric loading (about the x-axis), we can use
Equation 5.95 to obtain the contact pressuresat the two edges of the rectangle (Figure 5.35).

-
P P - 6el L) -\177 x
qmax \177_(1 + 6el L) -\177 x = Ll2,
qrrun
= -:--(1 = -LI 2 (5.98)
LB LB

where xo is replaced by give the familiar

e to expressions used in soil mechanics-geotechnical
engineering textbooks. the purpose of the calculation
For of shear forces and bending mo-
ments along the footing, the contact pressure at every point on the centre line is multiplied by
B to give the soil reaction (in kN/m) along the assumed one-dimensional beam.

Example 5.24
For the rigid rectangular footing shown in Figure 5.36(a), determine the contact pressure dis-
tribution and the position of the R-ZPL. Find the length of the footing required to achieve a
uniform contact pressureif its length can be increased only beyond the right-hand column.
Solution:
Take moments about the left edge to find the position of the resultant of the loading system:
0,2 x 500.0+ 5.2 x 1100.0 + 580 - (500.0+ 1100) x' = 0.0, x' = 4.0 m.
The resultant is a vertical load of 1600kN that acts 4 m from the left edge of the footing. It is
clear that the integration of the contact pressure is also equal to 1600kN and acts at the same

'

Figure 5.35. Linear contact pressure under a rectangular rigid footing)))

202 Soil Mechanics: Basic Concepts and Engineering Applications

5.0 rn
\177
[\177 ii

580 kN-m

\177500 kN \1771100 kN

\1770.4m \1770.4m
L = 5.4 m, B=2.0 m

(a)

Figure 5.36. Example 5.24.

point but in reverse direction.

e = x'- L/2 = 4.0-5.4/2 = 1.3 m. Using Equations 5.98:
= 1600.0 x
qrr\177,\177
1.3 / 5.4)/(5.4 x 2,0) - 362.1
(i + 6 \303\227 kPa (right edge).

qmin
= 1600.0\303\227(1
-
6\303\2271.3/5.4)/(5,4\303\227 2.0)
= -65.8 kPa (left edge).
The soil may not sustain the tensile stress at the left edge, therefore the zero pressure line
moves towards the right and qrnax increases to satisfy equilibrium:
Effective length of the footing or the distance of zero pressure Iine from the right edge:
=3(5.4/2-1.3) =4.2 m.

The maximum soil pressure from Equation 5.97 is:

= - - 380.9kPa,
qma.x (2\303\2271600)/[3\303\227 2.0(2.7 [.3)] (Figure 5.36(b)).
For a uniform soil pressure,the centroid of the footing and the point of application of the re-
suttant must coincide: L = 2 x x' = 2 x 4,0 = B.O m, q = 1600.0/ (8.0x 2.0) =lO0.O kPa.

5.3.2 Beam on an elastic foundation

The concept of a beam on an elastic foundation provides a general solution in the evaluation of
the contact and settlement under
pressure a flexible tboting, as it takes into account the elastic
propertiesof both the footing and the soil. In this regard the Winkler model (after Winkler in
1867) has given a satisfactory solution for soil-footing interaction as well as for retaining

structures such as sheet piles.An early model of Winkler foundation assumed the soil under
the footing to be a bed of parallel and discontinuous springs as shown in Figure 5.37. The
modulus of subgrade reaction is defined as:

k, =q/Se (5.99)

where q is the under a loading plate and Se is the corresponding settlement

contact pressure
(Figure 5.38(a)).This m kN/m . A typical
modulus stress-setdement
\177smeasured behaviour is
shown in Figure 5.38(b). The value of ks depends on the magnitude of q and Se and decreases
as both q and Se increase: thus the spring, model is not linear. The relationship between q and
5e depends also on tl\177e size of the contact area, and the results obtained from a 0.3 m square
loading plate cannot be readily related to a loading plate of 1 m by 1 m or to real footings. The)))
 Stress Distribution and Settlement in Soils 203

(b)

Figure 5.37. Winkler model.

relationship between ks and Modulus of Elasticity & has been extensively investigated, and
the proposal by Vesid (t961a and 1961b) is often used in design:

=0.65xi \177
\177 Es
k\177 (5.100)
i- g2
\"\177Efff

where Es, E.r are the Moduli of of soil and footing respectively, B is the width
Elasticity of the
footing, If is the second moment of area of the footing section, p, is Poisson's ratio for the soil,
and k's =ks. B. A simplified version of Equation 5.100 (basedon Equation 5.54) is also used:

- E\177s
k
s (5.101)
B(1-\177t 2)
The Winkler model can be simplified by replacing the bed of springs with a finite number of
springs of stiffness K where:

K=contactareacorrespondingtoasinglespring\303\227ks=B\303\227a\303\227ks=k'sa (5.102)
where a is the horizontal distance between the parallel springs. Consider now the beam shown
in Figure 5.39. Each guided spring support allows (only) for vertical deformation of the beam
(zi) equal to R
i
/ K, where Ri is the reactive force at each support.

\177 Se2

(b)

Figure 5.38. Determination of modulus of subgrade reaction by plate loading test.)))

204 Soil Mecfvanics: Basic Concepts and Engineering Applications

Figure 5.39. Finite beam on parallel springs.

Ri = = = Kz (5.103)
k\177.az i ksBaz i i

The probleln can be solvedusing traditional as well numerical methods such as finite
methods
differences and finite elements. In method,
the latter the segments connecting the springs may
be selected as simple beam elements.In this simplified Winkter model, the modulus of sub-
grade reaction ks is assumed to be constant along the length of the beam. Terzaghi (1955) sug-
gested that the error arising from this assumption may be neglected in most practical problems.
UIrich (1991 and 1994) suggested a method based on tbe theory of elasticity that utilises a

spring constant that varies along the length of the beam. The ACI (1988) proposed a design
procedurethat uses a Boussinesq-type verticaI stress distribution as the basis for the variation
of k_\177.A practical design procedure for plane strain conditions was proposed by Liao (1995)
who used finite elements as a replacement for the mathematical treatment of the theory of elas-

ticity. He obtained correction factors for ks that depend upon the position of the nodal spring
and the depth of the incompressible stratum.
A closed form solution for a beam of simpIe geometry and loading resting on a bed of
springs has been derived by Hetenyi (1946). Whilst the solution for an applied concentrated
moment and vertical toad on infinite and finite beams can be found in many geotechnical en-
gineering textbooks, there has been no practical development of the equations for distributed

loads and complicated geometry. Moreover, the flexuraf stiffness of the beam may vary along
its Iength, and in this case the derivation of a closed form solution is abnost impossible. The
. th
to the n - 2 degree,where
\342\200\242
beam shown in Figure 5.39 is mdeterm, nate n is the number of

springs. A traditional approach to the solution of this beam establishes supplementary equa-
tions based on the geometry of its deformations. The governing differential equation for the
deflection of ti]e beam is:

d2z _ - M (x)
(5.104)
dx 2 Et

where M(x) is the bending moment at distance x from the origin (say left edge) and El is the
flexuraf stiffness of the beam. Note that by accepting this differential equation, (vertical) de-
flections due to shear are ig,\177ored (an Euler beam rather than a Timosbenko beam). A practical
way of handling a beam on an elastic foundation is to use finite elements with constant spring)))
 Stress Distribution and Settlement in Soils 205

stiffness. An alternative method is to approximate the second derivative in Equation 5.104

with finite differences (Equation 3.52), an approach demonstrated in the following example.

Example 5.25

Predict the soil settlement and reaction at selected points for the finite beam shown in Figure
5.40. The thickness of the footing is 0.7 m and its width is l m. The beam is divided into 5
a
equal segments of 2 m each. El= 22 x 10 kPa, Es = 10440kPa, g = 0.3.
Solution:
Using Equation 5.100 to calculate k's:

10440x1\"04 10440
k\177
=0.65 l\177J.
3
=5300 kPa,
22x106(0.7) xl.0/12 1-0.32

K= k'sx a = 5300x 2.0 = 10600 \"kN/m. For the edges: K = k'\177x (a / 2) = 5300 kN/m,

Although (1996) recommended that the end stiffness be doubled, the model expressed
Bowles
by Equation 5.103 will be followed. Thus, R! = 5300 zl, R2 = 10600 z2, R3 = 10600 z3. Be-
cause of the symmetrical loading, the required parameters at points i, 2 and 3 are equal to the
corresponding values at points 6.5, and 4. To solve the problem three linear equations involv-
ing the three unknowns of zt. z2 and z,3 (settlements at points 1.2 and 3) must be developed,
Applying the finite difference equations at points 2 and 3, we have:

d2,- + -M
(- ,)i = Zi+ 1 Zi_
2
1 -2Zi _ i
'
a El

d2z - z 3 - -
2z'\177 -M 2
(7q-)2
+_ z_.t Z
'
2.02 El

d
2 .
+ Z2 - 2Z - Z3
- M
3 _ Z2 3
(- = Z,\177 _
Note z3 = z4.
-
\177)3
2.02 2.02 El

El -- 22.0x 106 x (l.0x0.73 /12)= 628833.33 kN.m 2.

Calculate bending moments at points 2 and 3:
M2= RI x 2.0 - 1000.0 x 1.0= 10600z\177 - 1000.0,
M3 = RI \303\227
4.0 + R2 x 2.0 - 1000x 3.0= 21200z\177 + 21200z2- 3000.0,
Substituting M2 and M3 into the finite difference equations:
z3 + zI - 2z
2
-
_ -(10600q 1000.0)
4.0 628833.33
-
\342\200\242 -(21200z I + 21200z - 3000.0)
:2 z3 _ 2
4.0 628833.33)))
206 Soil Mechanics: Basic Concepts and Engineering Applications

lO.Om

1000 kN 1000 kN

Figure 5.40. Example 5.25.

The third equation is obtained from (static) equilibrium of the vertical forces:
2Rt + 2R2 + 2R3= 1000.0 + 1000.0 =
2000.0,

RI + R2+ R3- [000.0or:

5300 zl + 10600z2 + 10600z3 = [000.0.
Simplifying and rearranging the equations:
671233.33zl - 1257666.66z2
+ 628833.33z3 = 4000.0

84800zl + 7 [3633.33z2 - 628833.33z3=

12000.0
5300z\177 + 10600z2+ 106007_3= 1000.0
Solving for zl, z2, and z3:
-3 -3 = 259.3kN
zl = 48.92 x 10 m = 48.9 ram, RI = 5300 x 48.92 x [0 $

_3 -3 =
z2 = 38.58 x 10 m = 38.6 rrLm, R2 = 10600 x 38.58 x t0 408.9\"kN$
-3 _3
z3 = 31.30 x [0 m = 31.3 mm, R3= 10600 x 31.30x 10 = 331.8 kN$

5.6 PROBLEMS

5.1 A cylinder 150 \177rLm and height 300 rnrn is filled with sand. The surface of
of diameter
the sand to a vertical stress of 300 kPa causing 4 mm settlement under
is subjected the

loading plate. Calculate the lateral stress on the wall of the cylinder and Modulus of
Elasticity of the sand. Poisson's ratio for the sand is 0.2.

Answers: 75 kPa, 20250kPa

5.2 Referring to Figure 5.5(a), calculate the distribution of vertical stress along a vertical

plane passing through one of the two forces. Specify the values of stressat depths of 1
m, 2 m, 3 m, 4 mand5 m.

Answers: 191.0 \"kPa, 48.1 kPa, 22.0 kPa, 13.1 'kPa, 9.0 kPa

5.3 Figure 5.41(a) shows a plan view of a footing that applies a uniform load of 300 kPa on

a horizontal grou,\177d surface. Calculate the vertical stress component at point A at a depth
of 2 m using:)))
 Stress Distribution and Settlement in Soils 207

6.0 m
m

\1773.0m \177I

(a) (b)

Figure Problems
5\303\26741. 5.3 & 5.5.

(a) The principle of superposition of vertical forces by dividing the area into elemental 1
m squares, (b) Newmark's influence chart.
Answers: 186.4 kPa, 184.6kPa
5.4 Re-work Problem 5.3 using Fadum's chart.

Answer: 184.8 kPa

5.5 Figure 5.41(b) a plan view of a rectangular
shows footing 2 m square hole
with a 2 m \303\227
(through thickness). The hole is locatedat I m frown the left edge and is equally
its entire

positioued between the top and lower edges of the tooting. If the uniform bearing pres-
sure under the tooting is 200 kPa, use Faduln's chart to compute the vertical stress com-
ponent at a point 2 m below the centre of the square hole.

Answer: 79.6 kPa

5.6 Compute or: at pointA (Figure 5.42(a)) due to the two line loads shown.
Answers: \177r
= 1.5 kPa, %:= 5.8 kPa, \"rxz= 2,0 kPa

5.7 Compute the magnitudes of the major and minor principal stresses within a soil that is
subjected to an infinitely long line load applied at the ground surf'ace.
Answers: \1773
= 0.0, : (2q /
\177r\177 :zz)cos2{3
5.8 Under the centre line of an infinite strip footing, calculate the depth at which the vertical

stress component is 10% of the applied load.

Answer: z - 6.34B
5.9 loading of infinite
A strip length is shown in Figure 5A2(b). Compute the vertical stress

component at points A, B, C, and D.

Answers: 81,8 \"kPa, 63.7 kPa, 14.8 kPa, and 2.8 kPa)))
208 Soil Mechanics: Basic Concepts and Engineering Applications

1.5m \177
\177i_\177

20.0 10.0 kN/m

i
kN/m\177

z =1.0 rn
A B C D - --
m i ..... \177 .... \"-,t ..... '-,I'-
1.0 rn 1,0 m 1.0 m

z
(b)
FiEare 5.42. Problems 5.6 & 5.9.

5.10 An earth embankment is 2.5 m high and has a slope of 2 horizontal to 1 vertical. If the
base of the embankment is 20 m, find the vertical stress component at a point on the cen-
tre line at a depth of 7
Answer: 74.5 kPa

5.11 A vertical pile carrying a load of 1500 kN has been driven 18 m into the ground. Calcu-
late the vertical stress component at a point 19.8 rn below the ground surface and 3 m
from the centre line of the pile for the following cases; (a) the entire load is transmitted
to the soil through the base of the pile, (b) the base carries one half of the load and the
rest is carried skin friction. = 0.3.
by \177.t

Answers: 7.9 kPa, 5.7 kPa

5.i2 Two flexible square are constructed 4 m apart (centre to centre)
footings 2 m \303\227
2 m
the ground surface. contact
The uniform
pressureunder one footing is 200 kPa and is
400 kPa under the other, Calculate the elastic settlement at the centre of each footing and
at the midpoint of the line connecting the two centres. E\177-- 10000 kPa, Ix= 0.35.
Answers: 51.0 mm (q = 200 kPa), 84.5 \177rma (q = 400 kPa), 33.7 trim
5.i3 A flexible circular footing 4 m in diameter exerts a pressure300 kPa at the ground sur-
face. The soil is a saturated clay with Es = 8000 kPa, and an incompressible stratum is
located \1772 m below the ground surface. Calculatethe settlement under the centre of the
footing using: (a) a semi-nu\177nerical method by dividing the depth into 6 layers of 2 m
thickness, (b) Steinbrenner's i\177fluence factors.

Answers: 100,0 ram, 76.4 mm

5.14 A flexible rectangular footing for which B - 2 m and L =4 m exerts 300 kPa 2 m below
the ground surface. The Modulus of Elasticity for the soil is 8000 kPa, and IX = 0.5, An
incompressible stratum is located at 1.0m below the ground surface, Calculate the aver-
age settlement of the footing using (a) Steinbrenner's influence factors and Fox's correc-
tion factor for deptl\177, (b) the improved chart of Janbu et at.)))
 Stress Distribution and Settlement in Soils 209

Answers: 39.4 ram, 59.5 mm

5.15 A footing has plan dimensions of 3 m and exerts a uniform
5 m
\303\227 net contact pressure of
200 kPa on the underlying satur;\177ted sand. base of the footing
The is at the ground sur-

face, and the average SPTnumber at a depth of 3 m is 19. Calculate the settlement of the
footing using Burland and Burbidge method.
Answer: i7.3 mm

5.16 A rectangular footing of width 2 t'n length of 20 m is located at the ground surface. It
and
exerts a net uniform contact pressure of 180 kPa to the underlying deep deposit of sand.
Calculatethe settlement of the footing using Schmertmann's modified strain influence
factor diagrams, The average results of CPTtests are shown in the table below.

Depth (m) 1.0 2.0 3.0 4.0 5.0 6.0 7.0

qc (kPa) 1600 1600 2000 1800 2400 2400 2600

For calculation of E,,. use the plane strain condition expressed by Equation 5,91.
Answer: 74.7 mm

5.17 Calculate the contact pressures under the comers of a rigid square footing using the fol-
data: P = 600 kN, Mx =
- 60 kN.m, = i00 kN.m, L = B = 2 m.
lowing My

Answers: 270.0 kPa, 120.0 kPa, 30.0kPa. 180.0kPa

5.18 Calculate the contact pressure for the rigid footing of Figure 5.43(a) at points a, b and c.

Answers: 51.3 kPa, 278.4 kPa, 391.8kPa

5.19 The data apply to a square footing located at the ground
following surface:
L=B= t.Sm, P=270kN, Mx=- 160kN.m, My = 160 kN.m.
Determine (a) the contact pressure at the corners of the footing, (b) the equation of the
zero pressure line, its readjusted position (by tbrmulatio\177x) and the maximum contact
pressure (assumereadjustme\177tt occurs parallel to the original zero pressure line), (c) the

size of the footing to limit the maximum contact pressure to 280 kPa and the position of
the new zero pressureline.
Answers: (a) 412.5, 83.3, 83.3, and -245.9 all in kPa,
(b) x + y + 0.4556 = 0.0, h I = distance from the maximum contact pressurepoint =

0.85 m, qmax =1125 kPa,

(c) 2.4 m, 1.7 m from the maximum contact pressure point.
5.20 For a trapezoidal rigid lboting shown in Figure 5.43(b) compute the dimensions B and A
that witt ensure a uniform contact pressure of 200 kPa.

Answers: B = 3.92m, A = 2.08 m

5.21 Re-work Example 5.25 by replacing the two vertical tbrces by a single vertical force of
2000kN applied at the centroid of the footing.

Answers: zl = 26.2 ram, z2= 36.4 ram, z3 = 44.8 ram,

kN, R3 = 475.4kN)))
-
R\177 i38.8 kiN, R2= 385.8
210 Soil Mechanics: Basic Concepts and Engineering Applications

(a) (b)

Figure 5.43. Problems 5.18 & 5.20.

5.7 REFERENCES

ACI, I988. Suggested analysis and design procedures for combined footings and mats, ACf structt\177ral

jo,rnal, 85(3): 304-324.

Ahlvin, R.G. & Ulery, H.H.1962. Tabulated values for determining the complete pattern of stresses,
strains, and deflections beneath a uniform load on a homogeneous half space. Highway research
board btdletin (342): I-13.
Atkinson, J. H., & Bransby, P. L. 1978. The mechanics of soils. UK: McGraw-Hill.
Bjerrum, L. & Eggstad, A. 1963. Interpretation of loading test on sand. Proc, _(tJ Eurpe\177n conf. on soil
rnechanics I: 199-203. Wiesbaden.
th
Bowles, J.E 1996.Foundation analysis and design. 5 edit.ion. New York: Mc Graw-Hill,
Burland, ].B. 1970. Discussion. Proc. conf on b\177-si\177in\177,es\177igations #7 soils and rocks. London: British
Geotechnical Society.
Burland, J.B. & Burbidge M.C. 1985. Settlements of foundations on sands and gravel Proc.. Pan 1, 78:
1325-138 I. UK: Institution of Civil Engineers.
Christian, J.T. & Carrier W.D. 1978. ]anbu, Bjermm, \177nd Kjaernsli's chart reinterprcted, Canadian
geotechnicaljournal 15(1): 123,436.
nd
Clayton, C.R.I., Matthews, M.C., & Simons, N.E. 1995. Site #tvestigation. 2 edition. London:
Blackwell Science.
Consoli, N.C. Schnaid. Milititsky, J. 1998.Interpretation
F. & of plate load tests on residual soil site,
Joun\177al GGE, ASCE 124(9): 857-867.
D'Appolonia, D.J, Poulos H.G. & Ladd C.C. 1971. Initial settlements of structures on clay. Jo,rnal
SMFEV, ASCE 97(SM\177 0): 1359-1377.
Duncan, J.M. & Buchigna\177ti, A.L. 1976. A\177tengb\177eering manual for settlement studies. Berkeley: Uni-
versity of California,)))
 Stress Distribution and Settlement in Soils 211

\"a
Fadum, R.E. 1948.Influence values for estimating stresses in elastic foundations. Proc. 2 intern, conf
SMFE 3: 77-84.Rotterdam.
Fox, E.N. 1948. The mean elastic setdement of a uniformly loaded area at a depth below the ground sur-
nd
face. Proc. 2 intern, conf SMFE 1: 129-132.Rotterdam
Geddes, J.D. 1966. Stresses in foundation soils due to vertical subsurface loading. Geotechniquel6(3):
231-255.
Geddes, I,D. 1969. Boussinesq based approximations to the vertical stresses caused by pile type
subsurface loadings. Geotechnique 19(4): 509-5 IZl-.

Gibson, R.E. 1974. The analytical method in soil mechanics: 14'\177'Rankine lecture. Geotechnique 24(2):
[ 15-140.
Giroud, J.P. 1972. SoAtlement of rectangular foundation on soil layer. Jou\177ml SMFED, ASCE 98(SM1):
149-154.
Hart, M.E. 1966. Foundations oftheoretlcal soil mechanics. New York: McGraw-Hill.
Hart, M.E. 1977.Mechat\177ics of particulatq media: A probbabilfstic approach. New York: McGraw-Hill.
Hetenyi, M. 1946, Beam on elastic foundations. Ann Arbor, Michgan: The University of Michigan Press.
Holtz, A.D.1991. Stress distribution and settlement of shallow foundations. In H.Y. Fang (ed.)
Foundation engineering handbook. New York: Van Nostrad Reinhold.
lrles, R. & lrles F. 1994. Explicit stresses under rectangular footings, Journal SMFED, ASCE 120(2):
444-450_
Janbu, N., Bjermm\177 Kjaemsli, B. 1956.Veileduing
L,, & red losning, av fundamenteringsopogaver (Soil
mechanics applied to some engineering problems).Noewegiangeotechuical insritiute Pub. 16. Oslo.
Jumikis, A.R. 1969. Theoretical soil mechanics. New York: Van Nostrand Reinhold,
Liao, S.S.C. 1995.Estimating the coefficient of subgrade reaction for plane strain conditions. Proc.
&st#ution of civil engineering, geotectdcal et\177gineering 113:166-18 \177.

Mayne\177 RW, & Poulos, H.G. 1999. Approximate displacement influence factors for etastic shallow
foundations, Jour\177,al GE, ASCE 129(6): 453-460.
nd
Meyerhof, G.G. & Fellenius, B.H. 1985,Canadian foundation engineering manual 2 edition. Canada:
Canadian Geotechnica[ Society.
Mindlin. RD. 1936. Force at a point in the interior of a semi-infinite solid. Journal of the American insti-
tution of physics 7(5): 195-202.
Newmark, N.M. 1942. Influence charts for computations of stresses in elastic soils. Enginnering
experimental station bulletitl (367). University of Illinois.

Papadopoulos, B.P. 1992. Settlements of shallow foundations on cohesionlcss soils, Jour\177u\177l SMFED,
ASCE 118( 3): 377-393.
nd
Peck, R.B., Hanson, W.E. & Thombum. T.H. 1974.Foundation engineerit\177g, 2 edition_ New York:
John Wiley & Sons.
Poulos\177 HG, & Davis, E.A. 1974.Elastic solutions for soil at\177d rock mechanics. New York: Wiley.
Powrie, W. t997. Soil mechanics-concepts a\177d applications_ London: E & FN Spon.
Schmertmann, J.H. 1970. Static cone to compute static settlement over sand. Jout7tal SMFED, ASCE
96(SM3): 1011-1043,
Schmertmann, ].H., Hartman, J.P. & Brown, P.R. 1978.hnproved strain influence factor diagrams. Jour-
nnl GE, ASCE 104(8): 113t-t 135.
Steinhrenner, W, 1934. Tafeln zur Setzungsberechnung. Die strassel.
Sutherland, S,B 1075. Granular materials. Proc, cot\177 o\177 settlements of structures: 473-499. BGS.
Cambridge. London: Pentech Press.
Terzaghi. K. 1955.Evaluation of coefficients of subgrade reaction, Geotechnique 5(4): 297-326.)))
212 Soil Mechanics.\" Basic Concepts and Engineering Applications

Terzaghi, K. & Peck, R. B., 1948,Soil mechanics in engineering practice. New York: John Wiley &
Sons.
rd
Ter-zaghi, K., Peck, R. B., & Mesri G. 1996. Soil mechanics in engineering practice. 3 edition. New
York: John Wiley & Sons.
r\177l
Timoshenko, S.P & Goodier. J.N. 1982. Theory of elasticity, 3 edition. 398-409. New York: McGraw-
Hill.
th
Tomilinson, M.J. 1995. Foundation design and edition. 6
Harlow, Essex: Longman
construction.
Scientific & Technical.
Ulrich, E.J., Jr. 1991. Subgrade reaction in mat foundation design. Concrete international 13(1): 41-50.
Ulrich, E.J., Jr. t994. Mat foundation design: A hisl\1773rical perspective. Proc. ASCE specialty conf. on
vertical and horizontal deformations of foundations and embankments. 107-120. Texas A&M,
College Station l-
#\177
Vesi\177, A.S. 1961a. Beams on elastic subgrade and the Winkler's hypothesis. Proc. 5 intern, conf on
soil mechanics and foundation e\177tgineering. 1 : 845-850.
Vesi\177, A.S, 1961b. Bending of beamsresting on isotropic elastic solid, Journal EMD, ASCE 87(2): 35-
53.
Wu, T.H. 1966. Soil mechanics. Boston: Allyn & Bacon Inc.)))
CHAPTER 6

One Dimensional Consolidation

6.1 INTRODUCTION

This chapter introduces methods to predict both the magnitude and rate of one-dimensional
consolidation settlement, Consolidationsettlement in a fully saturated clay is the result of a
gradual reduction in its volume due to the drainage of water through the voids of the soil, The
potential that causes this movement of the water is caused by the increase in the pore pressure
due to external Ioading. This increase in pore pressure within the soil is termed the excesspore

pressure and varies with time and from point to point within the soil mass. In a saturated clay
in the undrained condition, the applied load at time t = 0 is resisted solelyby the pore water. If
a drainage condition exists, the initial excess pore pressuredissipateswith time leaving the
soil, again, in a fully saturated state. Theconsequent reduction in the volume of the soil is ap-
proximately equal to the volume of water entering the free draining boundaries. The increase
in the effective stress at any giveq time equals the decrease in the excess pore pressure (be-
causeof the concept behind Equation 2A). This assumption, along with the continuity condi-
tions of water movement, enables the creation of the necessary equations that express the de-
velopment of the consolidation process. The settlement due to consolidation is the major
controlling factor in the design of footings and the like, and in many cases little error is intro-
duced by assumi\177tg consolidation only occurs in the vertical direction. Any immediate settle-
ment that occurs upon application of the load is estimated using elastic theory.

6.2 CONSOLIDATION INDICES AND SETTLEMENT PREDICTION

6.2.1 Assumptions and basic settlement model

The magnitude of the theoretical consolidation settlement is dependent on the level of the ap-
plied stress(applied by a footing for instance) and the adopted stress-settlement model that is

normally based on laboratory tests. Over the period 1925-35, soil mechanics pioneer Terzaghi
and his associate Frolich established the conventional consolidation theory that later became
one of the most widely used theories in geotechnical engineering (Terzaghi & Peck, 1967).
The limitations and shortcomings inherent in this theo\177 (Duncan, 1993) have been overcome
213)))
214 Soil Mechanics: Basic Concepts and Engineering Applications

by the application of more advancedmethods including modifications to the original basic as-
sumptions. In order to develop the equations necessary to predict the magnitude and rate of
one-dimensional consolidation the following assumptions are made:
1. The soil is homogeneous.
2. The soil is fully saturated.
3. Darcy's law is valid.
4. The flow is one-dimensional and the coefficient of permeability is constal\177t during the con-
solidation process.
5. The relationship between volume change and effective stress is independent of time but

changes with the level of stress.

6. Strains are small.
Figure 6.1(a) shows behaviour typical of an element of saturated clay undergoing one-
dimensional consolidation in the laboratory. Void ratio is plotted vertically against effective
vertical stress on a logarithmic scale. I1\177general, the void ratio decreases with an increase in
effective vertical stress. At small values of effectivevertical stress the rate of decrease of the
void ratio is corTespondingly low umil the effective vertical stress reachesthe preconsolida-
tion pressure P'c, The state of the soil at effective vertical stresses below p'c is called overco\177,-
sofidated. For effective vertical stresses greater than p'c, the e-log\177 curve can be approxi-
mated by a straight line, referred to as the virgin compression line, In this region the soil is
normally consolidated. Under one-dimensional consolidation the state of stress is either on the
virgin compression line or to the left-hand side of this line. No state of stress can be exist on
the right-hand side of the virgin compression line. An unloading-reloading cycle is also shown

Overconsolidated
\177
el) . .
Preconsohdahon
C \177[

Time

Reloading\177 irgin compression

\\ ,V.

(Rec\302\260m P res s

i\302\260n)\\line,/\177

P'c 1.0
\177
(b)
log ,5\"
(a)

Ftgure 6.1. (a) Void ra\177io-loga\" relauonship, (b) time settlement relationship.)))
 Ot\177e Dimensional Consolidation 215

in Figure 6.1 (a). As a result of unloading, minor expansion of the specimen occurs and a frac-
tion of the compression is recovered. The reloading or recompression segment is slightly
above the expansion curve, and approaches the normally consolidated state as the effective
vertical stress increases to the preconsotidation pressure where it was unloaded. It may be con-
cluded that the overconsolidated segment indicates a recompression stage after an unloading
stage in the field. The effective vertical stress at the point where the curve approachesthe vir-
gin compression line is the maximum effective vertical stress that has acted on the soil some-
time in the past and is termed the preconso\177idation pressure (p',,). The stress history for an
overconsolidated segment is better represented by the introduction of a dimensionless parame-
ter called the overconsolidation ratio (OCR). This is the ratio of the preconsolidation pressure
to the existing effective vertical stress on the element. In the field the OCR may be defined as
the ratio of the preconsotidation pressure to the existing overburden pressure.
Figure 6.1(b) shows a typical time-settlement relationship for an element of saturated clay
during a vertical load increment. The settlement (or the vertical compressive deformation) of
the element is divided into three segments, Si, So, and Ss. The first segment Si occurs immedi-
ately after application of the load and without any change in the amount of water in the ele-
ment: consequently, this is representative of the elastic settlement in one-dimensiona\177 condi-
tions. The second segment S\177. (A to B on the time-settlement curve) is called primary

consolidation, or in general, consofidation settlement. As time advances from A to B the ex-

cess pore pressure dissipates from its initial va\177ue to zero. The third segment S+ represents the
settlement under a constant effective vertical stress and is termed secondaryconsolidation.
The main reason for this settlement is unknown but is thought to be as a result of colloid-

chemical interactions and small residual excess porepressures. For inorganic soils the amount
of secondaryconsolidation is negligible.

6,2.2 Compression index and coefficient of volume compressibility

The compression index Cc is the absolute value of the gradient of the virgin compression line:

= e\302\260-el - e\302\260-el
Cc (6.1)
Iog\177 1 -Iogcr 0 log(\177s I/cry)

where the subscripl.s 0 and I represeat the initial and final states of void ratios and correspond-
ing effective vertical stresses, A similar equation can be to define
written the recompression
index Cr, which can be obtained from either the overconsolidated segment of the e-logo\" plot
or an unloading-reloading cycle, Leonards (1976)suggested that the unloading-reloading cycle
must be close to the preconsolida6on pressure,and that an average slope must be taken for ex*
pansion and recompression assuming that both occur o\1771the same line. By doing this it is as-
sumed that the slopes of all the recompression stages that could happen in the field, or that can
be performed in the laboratory, are approximately equal. If a laboratory specimen has been
subjected to distortion or compression during the preparation process, it is more convenient to

establish Cr from an unloading-reloading cycle.

For the secondary consolidatiom a compression index is defined by:)))
216 Soil Mechanics: Basic Concepts and Engineering Applications

= Ael (6.2)
Ce\177 Alogt
where Ae is the change in void ratio in the secondary consolidationstage, and Alogt repre-
sents one (log)cycleof time. The secondary consolidation can alsobe represented by the coef-
ficient of secondarycompression in the form:

Cas = Lk/-/o\177/ H 0 (6.3)

where AHa is the change in the specimen height (in a standard consolidation test) over one
(log) cycleof time and H0 is the initial height of the specimen.
The compressibility of a clay layer undergoing one-dimensional consolidation can also be
represented by the coefficient of volume compressibility mr which is the volume change per
unit volume per unit increase in effective vertical stress:

&/-//H\302\260
(6.4)
--
131 13o
where AH is the consolidation the initial thickness of the layer,
settlement, Ho is
and 13'o and

13'1are the average initial effective

and vertical
final stressstates
in the of the
layer. The mag-
nitude of rnv is obtained by a standard one-dimensional consolidation test for a stress level rep-
- 0%. In this case AH and H0 and initial thickness of
resented by 13'1 represent the settlement
th\177 soil specimen respectively. In a fully saturated soil undergoing a vertical deformation zXH,
the following can be derived from the phase diagram (Chapter 1):

l+e
Ae
- 0
(6.5)
A/-/ H
0
where e0 is the initial void ratio and Ae = el -eo, in which el is the final void ratio. Thus
Equation 6.4 can be written as:

AH/Ho_ I Ho-H \177- 1 eo-e 1

(6.6)
131-130 H 0 cr]-130 l+e0 13]-cr0
in which the units of my is the inverse of stress (m2/kN or m2/MN).

6.2.3 Discussion on compressibility indices and their correlation with basic soil properties

The magnitude of the compression index Ce in saturated clays varies from 0.1 to 0.5 depend-
ing on their plastic characteristics, and increases with increasing plasticity. In organic soils and
peat, the compression index may increase to 3, although in some cases (e.g. Mexico City clay)
it reaches as high as 10 (Mesri et al., 1975).Although the concept of consolidation is not ap-
plicable to sands, a value of Cc of about 0.05 is most likely to be expected for loose sands.In
overconsolidated clays the recompression i,tdex Cr is small compared to Co, but in the calcula-
tion of settlement it must nevertheless be taken into account. Typical values are within the

range 0.015 to 0.035 (Leonards,1976)and the index decreases with a decrease in plasticity.
Disturbed samples exhibit high values of recompression index. However,the unloading and
reloading cycles for void ratios tess than 0.42e0 ,nay give a reasonablevalue for recompres-)))
 One Dimensional Consolidation 217

sion and compression indices (Schmertmann, 1953). This means that the stress-settlement be-
haviour, plotted in the form of e-logcy',approachesthe field or undisturbed condition as the

stress level corresponding effectivevertical

passes the stress of the void ratio of 0.42e0. In
most engineering projects the compression index Cc may be assumed to be constant within

working stress levels. However, in instances where the stress level varies between low and
high values (such as airstrips) it is necessary to investigate the variation of Cc with depth and
stress level (Balasubramaniam & Brenner, 1981).
Secondary compression is quantified by the parameters Ca and 6.2 and 6.3).
C\177s. (Equations

Values of commonly
C\177_\177 vary from 0.0005 for an overconsolidated
clay to 0.1 for organic
soils andpeat. From the definition of both parameters it would appear that the secondary cont-
pression indices are dependent on both time and stress level. However, the ratio of Ca/ Cc

may be assumed constant in both the and recompression stages (Mesri

compression &

Godlewski, 1977; Mesri & Castro, 1987).This that, for the calculation of secondary
means
consolidation settlement, the magnitude of Ca can be assumed to be independent of the time
span of interest. Furthermore, it is convenient to assume that the parameter Ca in the field is
independent of the thickness of the soil layer and the load increment ratio (the ratio of the ap-

plied effective vertical stress to the effective overburden pressure) provided primary consoli-
dation occurs (Raymond, & Wahls, 1976).
The statistical treatment of plasticity and compressibility has been the subject of extensive
investigation. Numerous linear equations in which the compression index is expressed in
terms of the and some basic physical properties, such as the field
liquid
limit moisture content
w,\177and void ratio eo, have been proposed
initial (Table 6.1). Whilst these equations are very
useful, they are not universal. Rendon-Herrero (1980 and 1983) suggested that a universal re-
lationship between Ce and dry unit weight can be constructed by relating the slope of the zero-
air curve (Equation 1.17) to the slope of e-log\177 curve. Usually, a simple linear regression

equation with e0 can explain up to 80% of the variation of C\177 and Cr (Balasubramaniam &
Brenner, 1981). This percentageincreases if the applicability of the equation is narrowed.

Table 6.1. Empirical relationships between consolidation characteristics and index or physical properties.
Empirical equation Author
- 7)
C\177.= 0.007(LL Skempton (1944)
C, = 0.009(LL - 10) Terzaghi & Peck (1967)
Cc
= 0.37(eo + 0.003 LL + 0.0004
wn
- 0.3,1) Azzouz et al. (1976)
19 I67
\302\260\177
Cc = 0.48827(y\177./7d) Rendon - Herrero (1980)
C,. = 0.009 wn + 0.005 LL Koppula (1986)
Cc = - O.156+ 0.411eo + 0.00058 LL AI-Khafaji & Andersland (1992)
=
Cr 0.l 4(e0 + 0.007 ), C\177= 0.003 (w,, + 7) Azzouz et al. (1976)
Cr = O.O0(LL + 9)
Cr = 0.00566w\177- 0.037, C\177-= 0.00463 LL - 0.0l 3 Balasubramaniam & Brenner (198 I)
Cr= 0\17700463 LL Gs Nagaraj & Srinivasa Murthy (1985)
Ca/C,.= 0.032,0,025 < Ca< 0.1 Mesri & Godlewski (1977)
C\177/C,. to 0.07, Peats. organic
= 0.06 soils Mesri (1986)
Ca/C,.= 0.015\177o0.030. Sandy clays Mesri et al. (1990))))
218 Soil Mechanics: BasicCotlcepts and Engineering Applications

Example 6,1
In one-dimensional
a standard consolidation test the vertical load oll the sample was increased
from 107to 428 kPa and the consequent
kPa reduction in the thickness of the sample was
measured at 1,044 ram. The initial thickness of the sample (at 107 kPa) was 19 mm and the
void ratio was 0.841. CalculateCc and my assuming that the specimen is in the normally con-
solidated state. Using the Cc concept, calculate the total load on the sample to achieve a void
ratio ofO.71.

Solution:

From 6.5:
Ae 1+0.841
--\177e tllus Cc - 0,1011 : 0,1679.
Equation =0,1011,
.044 19.0 Iog(428.0/107.0)
- 1 0,1011 = 1.71x10-4 m2/kN.
Using Equation 6.6: m,,
i+0.841 428.0-107.0

of Ceby
0.841- 0.71- 0,1679-->
From definition Equation 6.1: o\177
= 645.1 kPa.
log(o i / 107,0)

6.2.4 Determination of preconsolidationpressure

Based on extensive laboratory testing, Casagrande (1936)proposedthe following procedure to
establish the preconsolidation pressure:
1. Plot the e-logo-' curve from laboratory resutts (Figure 6.2(a)).
2, Visually establish the point of maximum curvature M on the recom \177ression line.
3. From point M draw two lines MH parallel to the logo<axis and MT tangent to the curve.

\177 \177 C A
eo
e0

0.42 eo _oJ .......... []

,
log \177\" log o\"
P'o P'c

(a) (b)
\177
Figure 6,2. (a) Casagrande method to estimate prcconsolidationpressure, (b) in-situ e-log\177 curve.)))
 One D#nensionat Consolidation 219

4. Draw the line MV that equally subdivides the angle between these two lines and find the in-
tersection point of the virgin compression line with this line.
5. Draw a vertical Iine through the above intersection point to intersect the e-logo\" curve at the

preconsoIidation pressure P'c,

In generaI, the compressibility parameters obtained in the laboratory are assumed to repre-
sent those in the filed. However, it has been found that the slope of the virgin compression line
in the field is slightly greater than that in the laboratory. Schmertmann (1953) proposeda
method for the correction of the laboratory results (Figure 6.2(b)) as follows:
1.In the e-logo\" plot, establish point
A with coordinates ofp'c and e0 (initial void ratio).
2. Draw a horizontal line from 0.42e0 and find the intersection point of this line and the ex-

perimental curve B). The laboratory

(point and in-situ void ratios are equal at this point.
3. Connect A to obtain the in-situ
to B virgin compression line.
4. Calculatethe effective overburden pressure on the overconsolidated soiI (p'o) and locate this
pressure on the logo<axis. Draw a vertical line from this point and find the intersection point
of this line with a horizontal line drawn from e0 (point C).
5. From point C draw a line paraIlel to the average slope of the overconsolidated segment of
the experimental curve to obtain the overconsolidated segment in the field. The broken line
CDB represents the in-situ e-Iogo\" relationship.

6.2.5 One-d#nensional consolidation test

A diagrammatic section of a consolidation apparatus (oedometer), consisting of a consolida-
tion cell and a sample ring, is shown in Figure 6.3. The cylindrical soil specimen has a diame-
ter in the range 50 to 75 man and a height of 15 to 20 ram. There are also oedometerswith
thicker samples and larger diameters (Section3.3.2),but their use in standard testing is infre-

quent as it takes much longer any excess pore pressure to dissipate. The specimen is inside a
metaI ring that is held by the consolidation celI, and the thickness of the ring must be sufficient
to ensure zero radial deformation. This ring is either fixed to the base of the cell or is left to
float where it is supported by the friction developed on the periphery of the soil specimen. To
provide free draining boundaries, porous plates (stones) with a diameter slightly less than the
diameter of the ring are positioned on both faces of the specimen. The ring with the soil

specimen and porous plates is submerged in the water inside the consolidation cell. Load is
applied through a loading plate and the consoIidation settlement is measured by means of me-
chanical or electronicdevices to facilitate a plot of the time-settlement (or time-void ratio) re-
lationship. For each increment of load the void ratio is calculated after the settlement of the
is
specimen completed, which is normally within 24 hours. The time-settlement plot for each
load increment is used to obtain the coefficient of consolidation cv (Section 6.3.4) which in
turn is used to relate time to settlement (or time to degree of consolidation) in the field. It is
customary to double the load every 24 hours from an initial value of 6 kPa to 3200 kPa and,
for each iqcrement, a value of my is computed. However, an aIternative loading schedule, pos-
sibly including a rebound (unloading-reloading) cycle, may be more representative of field
conditions. Swelling that may occur in some clays under low stresses should be avoided and)))
220 Soil Mechanics: Basic Concepts and Engineering Applications

Load
,[ Loading plate
Water level T z\" Ring

\177
Porous stone Soil

Figure 6.3_ Diagrammatic view of a floaling oedomeler apparalus.

therefore the loading should be increasedto the next higher level. The initial effective vertical
stress may be set equal to the effective overburden pressure in the field for stiff
clays but
should be lower for soft clays. Referenceis made to standard test methods: ASTM D-2435-96,
BS 1377-6:1990
and AS 1289.6.6.1-1998.

Example 6.2
Data from a laboratory consolidation test is given in the first three rows of the table below:

Test points 1 2 3 4 5 6 7 8

(r'(kPa) 14.0 28.0 56.0 t12.0 224.0 448.0 896.0 0.0

Tolal AH(mm) 0.t00 0.170 0.495 0.860 1.390 2.t30 2.855 1.530
Void
r\177fio
0.7t4 0.708 0\177681 0.650 0.606 0.544 0.483 0.594
m\177.(m-/MN) 0.250 0_565 0.329 0.238 0.172 0.088

Plot the e-logo\" curve aqd calculate C(.and m,,. Gs = 2.70, H0 = 20.5 ,rim and w = 0.22.
Solution:
The final void ratio after unloading is calculated and substituted into Equation 6.5 to obtain the
initial void ratio at zero pressure:

e8 = wG s = 0.22x2.7=0.594, z\177e - l+e 8 - 1+0.594

) &e = 0.1286 (positive),
1.53 H8 20.5 -1.53
e0 = 0.594+0.1286= 0,7226.
Similarly, we can relate the initial condition to the condition at

the end of each load increment. Samplecalculation for test point 1:

(Ae/0.1) = (1+ 0.7226)
/ 20.5 --\177Ae = 0.0084, -
(negative), e| = 0.7226 0.0084= 0.7142.

The results are summarized in the fourth row of the table above axtd are shown in Figure 6.4.
For points 5 and 6: C c = (0.606-0.544)/log(44&0/224.0) = 0.206,

For points 6 and 7:C c

= (0.544-0.483)/1og(896.0/448.0) = 0.203.
Assume an average of 0.204.)))
 One Dimensional Consolidation 221
0.75
0.723
0.7
0.65
0.6
0.594
0.55
to\177.\1778=
C. 9
0.5

0.45

0,4
10 100 1000
log c\302\242

Figure 6.4. Example 6.2.

Samplecalculation for my for the stress range of 112to 224 kPa (last row of the table above):

1 0.650- 0.606
= 2.38x10-4,n2/kN = 0.238 m2/MN.
1+0.650 224,0-112.0

6.2.6 Computation of consolidationsettlement

Consolidation settlement is calculated based on either the Cc or m\177.concept. Consider Equation
6.5 and replace Ae with e0 - el, AH with Se and H0 with H (the initial thickness of a layer).
e
_ 0 -el
Sc H (6.7)
1+e0
(e0
- el) 6.1, we obtain:
Replacing using Equation

=\177c\177 (6.8)
s\177. \177og(,\177i/,\177;\177)
l+e 0

Similarly, by combining Equations 6.2 and 6.5, the secondary consolidation is given by:

Ss _aH
l+e
Alogt (6,9)
0

where e0 and H are the initial values at the beginning of the secondary consolidation, which
correspond to the final state of primary consolidation. The term A logt represents the time cy-
cle in the field (t / tprim\177)
and is often less than 10 (Terzagbi et al., 1996), This means that

Alogt may be taken as unity and is always smaller than 2.)))

222 Soil Mechanics: Basic Concepts and Engineering Applications

In using Equation 6.8 we note that both the initial and final states are located on the virgin

compression line. If the initial effective vertical stress o\"0 is tess than the preconsotidation
pressure (overconsoIidated state) and the final state is on the virgin co\177npression Iine (normally
consolidated), then the settlement should be calculated in two stages according to:

- CrH , , CcH , \342\200\242

Sc\342\200\242 tog(pe/a o) +--log(\177|/p e) (6.10)
l+e0 l+ep
where ep is the void ratio corresponding to the preconsolidation pressure. For a layered system
or when the compressibility characteristics changewith depth, Equation 6.8 becomes:
i=ll

Sc =
\177(Sc)i (6.11)

where n is the number of layers and (Sc)i represents the consolidation settlement of each layer.
Equation 6.11 can also be used for a soil with a constant Cc throughout its depth if a realistic
average incremental effective vertical stress for each layer is selected.
From the definition of my, we can find the consolidation settlement for a layer of thickness H:

Se = &t-I =
mull
-
(\302\242j](J'o) (6. t2)
The general form of this equation may be written as:

i I

where Ac\177| is the incremental effective vertical stress and Az is the thickness of each layer.
Example 6.3

The coordinates of compression tine are: o-'1 = 400 kPa, el = 0,80;

two points on the virgin
\1772
= 800 kPa, e2 = 0,75, In normally consolidated tayer of this soil is to be
tile field, a 3 m
subjected to construction works and the average effectivevertical stress will increase from 250
kPa to 450 kPa. Determine: (a) the consolidation settlement, (b) the load increment to cause a
25 mm final consolidation settlement and the corresponding void ratio.

Solution:
0.80-0.75 = 0,166.
(a) Calculationof C,. from Equation 6.1: C,. =
log(800.0/400.0)

Equation of the virgin compression line:

0.80-e - = 0.8-
Cc 0.166 -\177 e 0.166(1og\177'-log 400.0).
=Iog\302\2427'/400.0
- = 0.834.
Initial void ratio in the field: e0 0.8-0.166(log250.0-1og400.0)
S,, = [0.166x3000/(1 + 0.834)]1og(450.0/250.0) - 69.3 - 69 ram.

(b) S c = [0.i66x 3000/(1+ 0.834)] Iog[(250.0 + A\177r') / 250.0] = 25.0 ---\177zX\302\2427'

= 59.0 k_Pa.

Thus o-'| = 250.0 + 59.0 = 309.0kPa and e I = 0,8 -

0.166(1og309.0 - log 400.0)
= 0.819,)))
 One Dimensional Consolidation 223

6.2.7 Consolidationsettlement due to lowering of the water table

The drawdown of the water table in soft clays and silts introduces consolidation effects result-
ing in settlemeqt of the ground surface. This settlement is progressive and its magnitude and
rate are dependenton the thickness of the layer, the drainage conditions and the amount of

drawdown, The main reason for the consolidation is the decrease in the pore pressure that ul-
timately increases the effective ve\177ical stress approximately in proportion to the depth of the
drawdown. The magnitude of this type of consolidation settlement recorded around the world
has varied from 0,2 m up to 8 m ('Terzaghi et al., 1996)and may be predicted by conventional
methods. However, this predictiou in thick layers is unreliable due to uncertainties in the
drainage conditions and the variation with depth of the compressibility characteristics that are
not detected from somewhat disturbed samples,
Undesirablesideeffects of dewatering have been fully documented (Powers, 1985) together
with procedures to minimize the dangers of the resulting consolidation settlement to the main
construction project and qearby facilities. Drawdown of the water table may result from the
pumping and removal of water from the ground aquifer and construction site in water supply
projects and excavations for footings and the like. In the case of temporary retaining structures
such as sheet piles,the increase in effective vertical stresses results in the increase of the effec-
tive horizontal stress or effective soil pressurethat has to be compared with the reduction of
total water thrust due to the lowering of the water table. Zeevaert (1957)and Parsons (1959)
suggested that consolidation settlement could be preve.\177ted by injecting the water from the ex-
cavations into the backfill. However,an environmental study may be appropriate for this proc-
ess of removal and iqjection of water because of bacterial cm\177cerns. Despite the problems as-
sociated with drawdown of the water table, it is an effective technique to preconsolidate the
soil in order to minimize the effects of consolidationsettlement on a structure, which is con-
structed after the settlement has occurred.
Parry & Wroth (198 I) proposed a simplified analysis to estimate the overconsolidation ra-
tio due to dewatering. The effective vertical stress prior to lowering of the water table is calcu-
lated from = 0% + u0 where \17770is the total vertical stress and u0 is the pore
Equation 2.1:\17770

pressure. Similarly the effective vertical stress after lowering the water table by the amount of
\177 is expressed by: \177t = \177 + (u0
- AH?w) where ?\177 is the unit weight of water. The differ-
ence between the total verticat stresses prior to and after dewatering is due to the change in the

unit weight of the zone subjected to dewatering. Thus, the amount of increase in the effective
vemcal stress is:
- = Act' = - AH
cr\177 (r\177 AHyw (y.,.at -Ym) (6.14)
where Ym is the unit dewatered zone and Ysat is the saturated unit weight.
weight of the The

negative term in may be ignored

Equation 6.14if it is assumed that the unit weight of the de-

watered zone remains unchanged as a result of capillary action. If pumping is stopped after the
completion of the primary consolidation, the water level will recover and the overconsolida-
tion ratio is defined by:

OCR =\177- -_
% 1 \177
KHy,\177. -aYt(y\177at- -Tin)
(6,15)
\1770)))
224 Soil Mechanics: Basic Concepts and Engineering Applications

The magnitude of AH is controIled by the amount of consolidation settlement predicted for the
specified loads exerted by the structure. If the initial state is in overconsolidated range the

magnitude of AH Yw must be greater than P'c in order to increase the OCR to a required value,

Note that the increase in the effective verticaI stress is constant within the thickness.

ExampIe 6,4
A 4.5 normally consolidated layer of clay is overIain by a free draining sand of 3 m
m thick
thickness. cIay layer is resting
The on a very thick sand layer. The water table is at the ground
surface. Calculate the settlement of the ground surface if the water table is lowered 3 metres
by pumping. The compression index of the clay is 0.64 and a point on the field virgin com-

pression line has a void ratio of 0.823 corresponding to an effective vertical stress of 35 kPa.
= = 19.1 kN/m 3.
Ysand Yclay
SoIution:

The totaI and effective vertical stresses and pore pressures at the centre of the layer are calcu-
lated at the initial and finaI conditions:

c\177 =(45/2+3.0)\303\22719.1-(4.5/2+3.0)\303\2279.81=48,8 kPa.

= + - 2,25\303\2279.81 = 78.2 kPa.
\177] (4.5/2 3.0)\303\227I9\1771

Equation of the virgin compression line is:

e = 0.823- 0.64(Iog o-' - log 35.0), thus:
e 0 = 0.823- 0.64(Iog48.8 - \177og35.0) = 0.731.
Using Equation 6.8:
Sc = 0.64\303\2271og(78.2/48.8)\303\2274500/(1+0.731) = 341 mm.

Example 6.5
A strip footing is constructed on the ground surface of the soil described in Example 6.4. The
clay layer is divided into 3 layers of 1.5 m thickness. T he effective verticat stresses at the cen-
tre points of each layer (downward) have been increased by 103.2, 86.5 and 74.4 kPa respec-
tively. Calculate the final consoIidation settlement of the footing under the foIIowing condi-
tions: (a) The load is applied after completion of the pumping stage of ExampIe 6.4, (b) the
load is applied after the water table has returned to its original teveI at the ground surface

without affecting the void ratios (no expansionhas occurred), (c) the load is applied before
the lowering of the water tabIe.

Solutioa:
(a) The resuIts of calcuIations are summarized in the table below:

Layer (r'o(kPa) Ct (kPa) eo Sr (ram)

1 64.3 167.5 0.654 219

2 78.2 164.7 0.600 179
3 92.1 166.5 0.554 149
Totalconsolidalion settlement (ram) 547)))
 One Dimensional Consolidation 225

Void ratio Virgin compression line

Ex, 6.4

(Dewatering), Ex. 6.5 part

Ex. 6.5
part (b) 6.5 part
, (a)
IEx.

log
Figure 6.5. Examples 6.4 and 6.5.

The of layers due to dewatering

settlements are 140 ram, 114 ram and 97 mm (calculations not
shown); note that the sum of the above settlements does not match with the average answer
obtained in Example 6.4. Sample calculation for layer 1:
G0 = (0.75+ 3.0)\303\22719.1-0,75\303\2279.81 = 64.3 kPa.
= 64.3+103.2 = I67.5kPa.
cr\177
e0 = 0.823- 0.64(Iog64,3- log 35,0) = 0.654.
S,,= 0.64x log(167,5/64.3)(1500.0 - 140,0)/(1
+ 0.654) = 219ram,

(b) The unloading and reloading curves are equivalent line, as shown
to a horizontal in Figure
6.5. The initial state is represented by point D, and settlement during
there is no the recom-
pression from point D to point B. The consolidationsettlement is due to the virgin compres-
sion from B to E as tabulated below, Sample calculation for layer i is as follows:

0% (kPa} e0 \177't(kPa)
Layer Sc (ram}
Point B Point B Point E

1 64,3 0.654 138+0 175

2 78.2 0.600 135.3 132
3 92.1 0,554 137,1 1130

Totalconsolidation settlement(mm) 407

=
\177r\177) (before lowering or after recovering)
= (0.75 + 3.0) x 19.1- (0.75 3.0)
+ x 9.81 = 34.8 kPa;
similar calculations will yield 48.8 kPa and 62.7 k_Pa for layers 2 and 3 respectively.
= 34.8
\177r\177
+ 103.2 = 138.0kPa, Sc = 0.64x 1og(138.0/64.3)(1500.0 - 140.0)/1.654 = 175 ram.

(c) The results of calculations are tabulated below (see also Figure
6.5).
For layer 1 : S c = 0.64\303\227log0 38.0 / 34.8) x 1500.0/(I+ 0.825)= 315nun.)))
226 Soil Mechanics: Basic Concepts and Engineerit\177g Applications

o\"0 (kPa) (kPa

{r'\177
Layer Point A eo Point E
Sc (ram)

I 3\177,.8 0.825 138.0 315

2 48.8 0.731 I35.3 246
3 62,7 0.66I I37.I I96
Total consolidation settlement (mm) 757

6.3 SOLUTIONOF ONE DIMENSIONAL CONSOLIDATION

DIFFERENTIAL EQUATION

6.3.i Time and excess pore pressure relationships

The final settlement
consolidation theoretically takes an infinite amount of time to occur. Pre-
diction the magnitude of the consolidation settlement at any given time after the loading re-
quires the solution of the one-dimensional consolidation differential equation constructed on
the basis of assumptions described in Section 6.2.1. The continuity equation in one-
dimensional flow, where the volume of the soil decreases over time, is written as:

k \1772h,
\177
dxxdyxdz
AV
=\177, or
\177t
k .....
\1772h

2
AV /V

\177t
mvAfs\"

\177t
m
v
Ou e

8z 8z

Note that in one-dimensional volume change \177V/V ! H . Furthermore. the change in ef- = AH

fective vertical stress is equal to the change in the excess pore water pressure (Ao-' = Au). Ex-
pressingh in terms of the pore pressure(ue = y\177,.\303\227
h), the differential equation above becomes:
Ou
e - O2Ue
(6.16)
cv
Ot 2
OZ

The parameter cv is called the coefficient of consolidationand is defined by:

cv = k
lmv\177 w (6.17)

and is expressed in rnm2/min or m?/year. Consider a layer of thickness 2d with two free drain-

ing boundaries (FDB) at the top and the base of the layer (Figure 6.6(a)). Assuming that the

origin of the z-axis is locatedat the top free draining boundary, the following boundary condi-
tions apply:
(a) t = 0 and
0 < z <- 2d, Ue = ui (initial pore pressure equalsthe applied total vertical stress),
(b) t for
> 0,z = 0 and z = 2d, ue= 0,
(c) t > 0, for z = d, Du e/Oz = 0 (conditions of symmetry and location of the maximum ue).
The solution for the two-way drainage condition using a Fourier series is:
-
. ng.\177. \177Tt\177 /
(6.18)

Assuming a constant value for ui at t = 0 (Figure 6.6(a)), we have:)))

 One Dimensional Consotidatiol, 227

(a} (b) (c) (d)

Free draining boundary

= FDB
FDB d \177 Closed boundary = CB

(e) (f)

Figure 6.6. Different types of initial excess pore pressure distribution.

n.._\177 2,: ( . 'I/\177Z '1

\177' ( n2a2c\"t
-- 1
= (l-cosnn)lsm\177exl\177
u
e \177.\177 2 (6.19)
n\177 2d 4d
n=l \177 ) \"\177
A dimensionless variable called the time factor T,, is defined as:
= /d
2
T\177, cvr (6.20)
For even values of n. the magnitude of 1- cosngis zero, while for odd values it is equal to 2.
By assuming n = 2m + I and M = r\177(2m + I) / 2, Equation 6.20 becomes:

ue = \302\243 -T,,) (6.21)

,,,=ozui/sin\177
M
\177 )

6.3.2 Degree of consolidation

The degreeof consolidation is defined as:
U: = (u i
- u /u
e) i (6.22)

For a uniform u, at time 0. ue = ui and the degree of consolidation is zero.As t --\177\177, Ue --\1770
t =
and degee of consolidation approaches1.At time t, the degree of consolidationgives the in-
crease in effective vertical stress as a proportion of the initial excess pore pressure,which in-
dicates the progress of consolidatim]. Substituting ue from Equation 6.21 into Equation 6.22:

\"\177
U. = [- \177, --|sin [expt-\177v\177 2T v) (6.23))))
228 Soil Mechanics: Basic Concepts and Engineering Applications

Uz (%)

0 10 20 30 40 50 60 70 80 90 100
0
0.2
0.4
0.6
0,8
1
1.2
1.4
1.6
1,8
2
Figure 6.7. Consolidation isochrones.

Equation 6.23 may be presented graphically in a two-dimensional coordinate system as shown

in Figure 6.7. The degree of consolidation shown on the abscissa varies from 0 to t (or 0% to

[00%). The ordinate shows the relative depth parameter z / d that varies from 0 at the upper
surface of the layer to 2 at the lower surface of the layer. This results in a series of curves for
various (constant) values of T,.. These curves are termed Isochrones, Each isochrone shows the

deveIopment of consolidation within the soil mass at a specified time.

6.3.3 Characteristics of isochrones

Figure 6.8 shows an isochrone for a value of T\177= 0.4. At point B with a depth ratio of 0.5 the
degree of consolidation is AB = 66.44% which means that 66.44% of the initial excess pore
pressure created at this point is dissipated, This is also equal to the increase in effective verti-
cal stressat point remaining excess porepressureis CB = 33.56%of the initial value.
B, The
For a two-way drainage system the maximum excess porepressureis at the centre of the layer.
For Tv = 0.4 this value is 47.45%. At this point the degree of consolidationis a minimum that
is given by 100,0 - 47.45=52.55%. For a one-way drainage system, only the half of the iso-
chroneis used.Note that in this case the maximum drainage path d is equal to the depth of the
layer and point B represents the midpoint of the layer. The maximum excess pore pressureis
on the impermeable boundary and consequently points on this boundary will have the
minimum degree of consolidation.An average degree of consolidation for the layer can be de-
fined by dividing the area between the isochrone and the left-hand axis (side AB) b\177'the totaI
area of z/d \303\227 [. The area within the isochcrone, if divided by the total area, gives the remain-
ing average pore pressure (in terms of a ratio or percentage of the initial excess pore pressure).
For the example above (Tv = 0.4), it can be shown that the area betweenthe isochrone and the
z / d axis is 1.3958.Sincethe total area is 2, the average degree of consolidation for the layer is)))
 One Dimensional Consolidation 229

Uz (%)

0 10 20 30 40 50 60 70 80 90 100
0
0.2
0.4
0.6
................... .......
% = 66.44 % % = 33,561%
0.8 (1-Ue/Ul)
\177,#'-(ue/ui)

1
1.2
1.4
1,6
1.8
2

Figure 6.8. Characteristics of an isochrone.

therefore 1.3958 ! 2.0= 69,8%. If the Uz and z / d values are multiplied by ui / I00 and d re-

spectively to give Ue and z, then the area between the isochrone (plotted in the Ue-Z coordinate
system) and z-axis, when multiplied by m\177, will yield the settlement at that specific time
(based on Equation 6,12). Similarly, the total area (z x ui) multiplied by my will equal the final
settlement providing that m\177 is constant within the layer, Consequently, the area between two
isochrones represents the incremental settlement between two given times.
The slope at any point on the isochrone representsthe hydraulic gradient related to the
steady flow of water. For an isochrone plotted in the ue-z coordinate system, the hydraulic

gradie\177t is equal to (1 / y\177.)(3u e [3z).

The average degree of consolidation is expressed mathematically by:

2d 0
U= , or (6.24)
u
i

2d

ui(2d)- ;uedz
U- 2d 2d
ui (2d)

0 0
S
U= \177
(6,25)
Sc

where S is the consolidation settlement of the layer of thickness 2d at time t and Sc is the final
consolidation settlement. The degree of co\177sotidation can be expressed in terms of the time
factor T\177,by substituting the excess pore pressure(Equation 6.21) into Equation 6,24:)))
230 SoilMechanics:BasicConcepts and Engineering Applications

0 0.1 0,2 0,3 0.4 0.5 0.6 0.7 0.8 0.9 \177

0.\177

0.2

0.3

0.4

U 0.5
0.6
0.7
0.8
0.9

Figure 6.9. Relationship bctween averagedegree of consolidalion and lime factor.

C =1- y\177 (6.26)

,\177-\177-exp(-M2Tv)
m=0 M
This equation is plotted in Figure 6.9 (curve 1), and is approximated with reasonable accuracy
by the following:
2
Tv =nU ]4--\177U <0.6, Tv = -0.933log(l-U) -0.085--,',U >0,6 (6.27)

For a half-closed layer (for which only one boundary is fi'ee draining), Equation 6.26 - with d
the thickness of the - is used if the distribution of 1\177iis uniform within the
representing layer
layer (Figure 6.6(b)). Equation 6.26 is also valid for the case of a linear distribution of initial
excess pore pressure with two free draining boundaries, where at one boundary the initial ex-
cess pore pressure is zero and at the other is equal to ui (Figures 6.6(c) and 6.6(d)). If the dis-
tribution of initial pore pressure is linear, from zero at the open boundary to t\177iat the closed
boundary with uiz = zxu i/d (lZigure 6.6(e)), then the average degree of consolidationcan be
obtained by substituting this liqear distribution into Equation 6.24:

U = 1-4 y\177 2Tv) (6.28)

,--\177-(smM)exp(-M
m=0 M
If the distribution of initial pore pressure is linear, from tti at the open boundary to zero at the
closed boundary (Figure 6.6(f)), then the average degreeof consolidation is:

U = 1- 4 y\177 M)(M sin M - 1)exp(-M-Tv ) (6.291)

,--\177-(sm
m=0 M
In practice, sufficient accuracy is achieved if the first seven terms are used, i.e..m increases)))
 One Dimensional Consolidation 231

from 0 to 6, Equations 6.28 and 6.29 are shown in Figure 6.9 (curves 2 and 3 respectively).
For a non-linear initiaI excess pore pressure, the average degreeof consolidation may be cat-
cuIated from:
2d

uedz
- -ue 0
-1 -
ui
U
2\"\177 (6.30)
ui
z
;uid
0
Table 6.2 shows the U-Tv relationship for a parabolic distribution of initial excess pore pres-
sure in which both sides of the layer are free draining boundaries, and ui is the maximum ex-
cess pore pressure at the middIe of the layer,

Table 6.2. U and T,. values for parabofic initial excess pore pressure.
T\177. 0.000 0.048 0.090 0.1f5 0.207 0.28f 0.371 0.488 0.652 0.933 ,\177

U 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Example 6.6

A normally consolidated 4 m thick layer of clay lies between two strata of free draining coarse
sand.PIot the
correspo\177nding
isochrone and determine the average degree of consolidation after

6 months, my = 0.24 m\177/MN, cv= 2.8 m'/year.

Solution:
e =
From Equation 6.20, T v = 2.8\303\2270.5(year)/(2.0) 0.35.
Use Equation 6.23the degree of consolidation
to find at depths 0.0, 0.5, 1.0, 1.5,2.0,2.5,3.0,
3.5 and 4.0 metres.Due to symmetry, the calculation is carried out for 0.0 to 2.0 metres. The
corresponding reIative depths (z / d) are 0.0 0.25 0.50and 1.00. For values of m > 1, the
magnitude ofexp ( M'T\177,) is negliDble, and calculatmns are carriedout only for m = 0 and m
= 1. The resuIts are tabuIated below and shown in Figure 6.10.

sin( M z / d)
2
m M 2/ M 2/ M Exp (- M'T\177,) z/d
0.250 0.500 0.750 1.000

0 0.5 \177z 1._73- 0.8106 0.4216 0.3827 0.7071 0.9239 1.0000

I [.5 \177z 0.4244 0.0901 0,0004 0.9239 0.7071 - 0.3827 - 1.0000

z!d 0.250 0.500 0.750 1.000

x 10
-3
379.614 \303\227
-3
x [0
-3
u(ot/ui 205.445 10-3 495.990 10
\303\227 536.855

0.165 [0
\303\227
-3
0.[26 x [0
-3 - 0.068 x 10
-3 - 0.178x IO -3
u\177)/ui
-3 -3 -3 .3
Total ue/ui 205,610x 10 379\177740 \303\227
10 495.922 x 10 536.677 \303\227
10
- 0.7944 0.6203 0.5040
U: = I / ui
u,\177 0.4633)))
232 Soil Mechanics: Basic Concepts and Engineering Applications

0 20 40 60 80 100
0

0.25
0.5
0.75
1
1.25
1.5
1.75
2

Figure 6.10. Example 6.6.

Using Equation 6.26: U = 1 -(0.8106x0.4216 + 0.0901\303\2270.0004) = 0.6582 = 65.82%,

From Equation 6.27: T v = -0.933 log0 - U) - 0,085 = 0.35 --\177U = 0.6582 = 65.82%.

6.3.4 Determination of the coefficient of coosolidation Cv

The parameter Cv is determined on the assumption that a theoretical relationship accurately

represents the experimental relationship. Traditionally, two methods based on different time
bases have been used.

Squareroot timefitting In this method, settlements measured during

method. a load increment
are plotted against the square root time. Figure 6.11 shows the theoretical consolidation curve

(Equations 6.26 or 6.27) in which the variation of degree of consolidation is plotted against
square root of time factor. From this plot the following observations can be made. The theo-
retical curve represents a straight li,le up to U = 60%. From Equation 6.26 ((\177r 6.27) for U
= 90%, Tv = 0.848. This is 1.333 times greater than the time factor calculated from the linear
2
part of the curve: T v = \177U214 = r\177\303\227(0.9) 14 = 0.636 and therefore 0.848 / 0.636 = 1.333.
Consequently, a line drawn from the point of zero consolidation with a slope (measured from
the vertical axis) of \1771.333 = 1.15 times greater than the slope of tinear part of the curve, will
intersect the curve at
\177 of 90%consolidation. These observations lead to the following pro-
cedure for obtaining t90 and the calculation of cv from the experimental curve:
1. Plot the settlement against the square root time.
2. Find the intersection of the linear part of the plot with the vertical axis and hence establish
the position of zero consolidation settlement. This point should be below the experimental
zero settlement point indicating an immediate settlement just before the commencement of the
consolidation process. Note that in the theoretical curve these two points are represented by
one point, as the itmnediate settlement is not included in the theoretical equation.
3. From the point in (2) above, draw a line with a slope of 1.15 times of the slope of the linear)))
 One Dimensional Consolidation 233

0%

2
Tv=(:,-d4 )U

Tv = -0.933 log(1 - U)
- 0.085

1 unit \177
\177!\\
1.15 unit I
\177

Figure 6.11. Theoretical consolidation in the square roo| (T\177.)-U coordinate system,

portion of the plot (measured from the vertical axis). The intersection point of this line with

the experimental plot represents t\177-90.

4. Substitute t90 into Equation 6.20, noting that at 90% consolidation Tv= 0.848.

c v = 0.848d 2/tgo (6.31)

Equation 6.31 can aIso be arranged in the folIowing form:

cv = O.
1114H2av/tgo (6.32)

where tg0 is inminutes, Hay is the average thickness (ram)

= 2d of the specimen for the load
increment and cv \177sm m /year.
5. Establish the primary consoIidation or 100% level by proportion, increasing the 90% set-
tlement by ten-ninths.

log of thne fitting In this method, settlements are plotted

method. against Iog time. Figure 6.12
shows the theoretical curve, where the variation
consolidation in the degree of consolidationin
terms of time factor is plotted on a semi-/ogarithmic scale. While the earIy portion of the theo-
retical curve is parabolic,the middle portion is linear. The intersection of this line, represented

by AB, with the tangent to the third and lower part of the theoretical curve located on the hori-
zontal axis shows the position of \17700% consolidation. The foregoing observations lead to a
straightforward step-by-step procedure for obtaining ts0 and calculation of cv from experimen-
tal data:
1. Plot the settlement in terms of log time. Note that zero time corresponding to zero settle-
ment cannot be shown on a logarithmic scale.
2. Select a point on the curve close to the settlement axis and find the corresponding time t for
this point. Locate another point such that its time is 4t. The settlement between these two
points is equal to the settlement between times t / 4 and t. The point with t / 4 could be taken
as the start time if t is relatively small. Thus, the position of zero settlement can be found.)))
234 Soil Mechanics: Basic Concepts and Engineering Applications

U= 0%
\177 Tv (log scale)

_u_ ........................

Figure 6. ] 2. Theoretical consolidation in the log (T\177)-U coordinate system.

This point should be beneath the experimental zero settlement point indicating an in\177nediate

settlement just before the comlnencement of the consolidation process.

3. Draw a tangent portion of the plot. This line is either horizontal
to the lower or slightly in-
clined thereby a secondary consolidation process. Find
showing the intersection point of this

line with the linear portion of the plot and hence establish the settlement corresponding to
100% consolidation. Find the midpoint between 0% and 100% to determine t.so.
4, Substitute tso into Equation 6,20 noting that at 50% consolidation Tv= 0.197,
cv =0,197d
2/t50 (6.33)

Equation 6.33 can also be arranged in the following form:

=
c\177 0.0259Ha2v/tso (6.34)
where ts0 is in average thickness of the specimen
minutes, H\177v is the for the load increment in
mm and c,, is immediate settlement and secondary
in m2/year. If the consolidation do not con-
trol the consolidation behaviour, the average thickness and other parameters are calculated
basedon the initial and final conditions recorded in the test; otherwise, correction factorshave
to be applied according to the relevant codes.

Example 6.7
Data obtained from a laboratory consolidation test on a clay is given below.
\1770 kPa, \177 = 200 kPa, Gs = 2.72, H0 = (initial thickness at 100 kPa) = 18 ram,
= 100

w = (moisture content at the end of the test) = 0.245.Determine: (a) cv from the square root
2
\342\200\242
time plot in m2/year, (b) Cv from the log time plot m m /year, (c) the coefficient of permeabil-
ity, and (d) the compression index Co. Assume the specimen is normally consolidated.

Solution:

(a) & (b) H,v = 18.0- 1.03 / 2 = 17.485 ram, d = i7.485/ 2 = 8.74mm.
square root time plot is shown from which
The in Figure 6.i3. = 3.45 too
= 11.90 min.)))
t\17790
 One Dimensional Consolidation 235

Time Total AH Time Total AH Time Total AH

(min) (ram) Onin) (ram) (min) (mm)

0.25 0.160 6.25 0.611 49 0.866

0.50 0.208 9 0.694 64 0.878
1.00 0.276 16 0.789 81 0.894
2.25 0.391 25 0.834 100 0.905
4.00 0.507 36 0.850 300 0.954
1440 1.030

cv = 0.848\303\2278.742\303\22710
6/[11.90/(365\303\2271440)] = 2.86
me/year.
The log time plot is shown in Figure 6.14 frown which is0 = 2.7 rain.
cv = = 2.93
0,197\303\2278.742 \303\227I0
-6/[2.7/(365\303\2271440)] m2/year.
Gs = 0.245\303\227
- w
(c) Calculatem,,: el(final) 2.72 = 0.666. From Equation 6.5:
\177e/1,03 = (1 + Ae + 0.666)/18,0 ---> Ae = 0,101, e 0 = Ae + e] = 0,101+0.666 = 0.767.
- -4
From Equation 6.6: mv = [1/(1.0 + 0.767)] \303\2270.101/(200,0 100.0) = 5.72 \303\227
l0 me/kN.

Using Equation 6.17 with an average value of cv = 2.9

me/year:
k = c,,mvY W = 2.9\303\2275.72\303\22710 -4 \303\2279,81/(365\303\22724\303\22760\303\22760)
= 5.16\303\22710-]0m/s.

(d) C c = 0.336.
= 0,101/1og(200,0/100.0)

Example 6.8

Re-working Example 6.3 (with free determine: (a) the time in days for
drainiug boundaries),
10%, 20%.... 50% ..... and 90% (b) the time i.\177days for 25 mm of
consolidation in the field,
settlement to occur, (c) the settlement at the end of 3 months, In a laboratory consolidation test

on this soil, the time required for 50% consolidationof a 20 rran thick specimen was 7,5 rain.

#time
(min)
0 5 10 15 20
=3\"45
O, t.\17790

0.4 \"
I
\177

0.74'...=
0.8 f.. \177 I

].2

Figurc 6,13, Settlement versus square roo\177 time: Example 6.7.)))

236 Soil Mechanics: Basic Concepts and Engineering Applications

time (min)
1 10 I O0 1000 10000
t50=2.70
..... ', \177i .....
\342\200\242
\177-

IIIIIII

...... IIIIIII

\177 IIIIIII
'
\"\177\"\177'\"\177- IIIIIJl
...... \"\177
IIIIIII
...... 1111111

Figure 6.14. Settlement versus log time: Example 6.7.

Solution:
(a) cv = 0.197 xl02 xi0 -6/(7.5/1440) = 3.78x10-3m2/day.
The times required for 10%,
20%,.. and 90% of consolidation in the field are calculated and tabulated below. Sample cal-
50% consolidation: cv = 3.78 \303\227
-3 = 2
0.197 x (3.0/2) / ts0 = 117.3
culation for 10 --\177
ts0 days.
(b)U = 25.0mm/ total consolidation settlement = 25.0 / 69.3 = 0.3607= 36.07%,

U (%) 10 20 30 40 50 60 70 80 90
Tv 0.008 0.031 0.071 0.126 0.197 0.286 0.403 0.567 0.848
Time (days) 4.8 18.4 42.3 75.0 117.3 170.2 239.9 337.5 504.8

2 .3 = 0.I02x
Tv =
r\177(0.3607) /4 = 0.102, cv = 3.78 x 10 (3.0/2)
\177
/t --\177t = 60.7
days,
-3
(c) cv = 3.78x10 =T v x(3.0/2) 2/(3x30)--'\177 Tv =
0.I51,
Tv =O,151=TtU7/4--\177U =0.4385,S c =0.4385x69.3
= 30.4 rnm.

6.3,5 Vertical drains

To facilitate the horizontal flow of water and thereby accelerate the consolidation process
within a layer of saturated sof[ clay, vertical sand drains or wick em- drains are sometimes

ployed. The diameter of a sand drain varies between 160 mm and 610 ram. It is constructed
using either an auger or a mandrel (Landau, 1966). In both methods some disturbance to the
soil occurs around the perimeter of the drain, making a smear zone with lower permeability
characteristics. Wick drains are prefabricated from a geotextile to produce drains 100 mm to
300 mm wide by 4 to 6 mm thick. Whilst different types of prefabricated drains have different
characteristics (Rixner et al., 1986;Bergado et al., 1996), they normally consist of a plastic
core for vertical draining paths and a filter jacket to prevent the passage of fine particles into)))
 One Dimensional Consotidatfon 237

Membrane To vacuum pump

Embankment

drans blanket drains

[J'\177 \177\177 \177\177 \177\177 [\177 \177\177

(a) (b)

Figure 6.15. Applications of vertical drains,

the core. The characteristics include geometric shape, configuration, drainage capacity and
flow resistance, and its likely effect on the permeability of the adjacent soil (see case studies in
Pilot, Figure 6.1.5(a) shows the application
198f). of vertical drains in conjunction with a pre-
loading system in the form of a vacuum pump, as originally suggested by Kjellman (1952). In
a deep clay layer this technique may not be sufficient when used alone, and vertical drains are

designed either with preloading at the surface or under gradually increasing construction load
as shown in Figure 6.15(b). Drains are arranged in triangular or square patterns (Figure 6.16).
The influence zone of each drain defines a soil cylinder of diameter De that is a function of the
distance L between the drains:

De = 1.1_28L for rectangular pattern, De = 1.505 L for triangular pattern (6.35)

The differential equation of one-dimensional consolidation for radial draining towards a sand
or a wick drain is constructed using principles similar to those in Section 6.3.1.

1_
c (Ozue + Oue\" - Oue
(6.36\177
\177--7- 7 0--;- O,
\"[
where Ue is the excess pore pressuredue to radial drainage only, t is time, and r is the distauce
of any point from the vertical drain. The parameter ch is called the coefficient of consolidation
in the horizontal direction and is defined by:

% = k\177r/ m\177.y\177. (6.37)

where kt\177is the coefficient of permeability in the horizontal direction. The solution of Equation
6.36 was obtained by Ban-on (1948) after assuming uniform surface loading and uniform ver-
tical deformation together with the following: at t = 0, ue = ui; and at t = t, Ue = 0 in the drain.
The degree of consolidationof the layer was found to be:
= t - / F(n)] (6.38)
U\177, exp[-8r\177,

where n, F(n), a\177d T\177are defined below in which D,, is the diameter of the drain:

n = De / Dw (6.39)

=
Tt, Cht / D2e (6.40))))
238 SoilMechanics:
Basic Concepts and Engineering Applications

o o
To ',
.....
I I ck.

--o i
o o

o Io o Io

o Io I o Io

(a) (b)
Fig:0re 6.16. Two traditional patterns for vertical drains.

n2 3n2-1
F(n)-- 2 (6.41)
nZ ltn(n) 4n

For the values of n > 10 F(n) may be approximated by:

F(n) = In(n) - 3/4
(6.42)

Figure 6.17 shows the variation of the degree of consolidationwith the time factor and n. For
comparisonTerzaghi's one-dimensional consolidation (Equations 6.26 or 6.27) is also shown.
Hansbo (1979, 1981, and 1987) modified Equation 6,38 to include the effects of the smear
zone and drain resistance by replacing F(n) with a general parameter F given by:

F = F(n)+ Fs + Fr (6.43)
Fs = (kt,/kt, s
-
1)In(D.,. I D,,,), Fr -
= 7\302\242z(H
z)k\177.,./qw (6.44)

The diameter of the smear twice the diameter of the mandrel

zone Ds is taken used in the
to be
installation of the vertical its coefficient
drain, of permeability
whilst in the hori-
kt\177sdefines

zontal direction. In the absence of experimental data 17\177scould bc assumed to be equal to the
coefficient of permeability in the vertical direction. The parameters H and qw are the length
and discharge capacity of the vertical drain respectively, where the discharge capacity is the
flow that passes through the drain under a hydraulic gradient of 1.0.If drainage occurs only at
one end the length H must be doubled. The parameter Fr represents the drain resistance at any
point with a distance z from the drainage end and has the following average values for the two
possible drainage conditions at the ends of the drain:
Fr = 27\177H
13q,,
2kt\177,\177 (one end), Fr = \177H
2khs 16qw (both ends) (6.45)
The contribution of any vertical flow can be incorporated by the inclusion of the correspond-
ing differential equations:

c,, (6,46))))
+7
L
 Ot\177e Dimensional Consolidation 239

Th and Tv
0.001 0,01 0.1 1
0

0.3 \342\200\242

3 0.4-

o.5
1
.... \"
0.6 '
o.7
lO,..j
0.8 . \177
\\
0.9
1 '

Figure 6.17. Degree of consolidation versus time factor for a clay layer with vcrlical drains.

The combined excesspore pressure (Carillo, 1942) is given by:

ue = (6.47)
Uev XUeh/u i

where is the excess pore pressure due to vertical

Uev draiuage only (Terzaghi's solutio,0 and
u& excess pore pressuredue to radial flow (Barro\177fs solution).
is the Consequently, by apply-
ing the defiaition of average degree of consolidation in Equation 6.47, the combined degree of
coqsofidation can be written in the following form:

U = t - (1 - U\177,)(1
-
U\177.) (6.48)

where U\177,and U,. are the degrees of consolidation for radial and vertical flow respectively,

Example 6.9

A soft clay Iayer l0 m thick is resting on au impermeable boundary and is overlain by 3 m of

sand, whose upper boundary is the ground surface, The clay layer is in the normally consoli-

dated state and the water table is 2 m below the ground surface. A load orS0 kPa is applied at
the ground surface over a large area. If sand drains are constructed in square pattern with L = 3
m determine the consolidatkm settlement after 6 mond\177s and the time required for 90% coq-
'. . . 3 3
sohdauon. \"7 (sand above water table) = 18 kN/m , y,, (sand) = 20.2 kN/tn , g,.,\177
\" (clay)
= 19
3 2 2 =
kN/m, c\177,= 5.I5 m /year, ct\177 3._0 m /year, Cc 0.25,e0=0.8.
= 9
Solution:

Initial and final effective vertical stresses at the mid point of layer are:
=
cy\177) 19.0x 5.0 + 20.2x 1.0+ 18.0x2,0- 9.8Ix6,0=92.3
kPa,

c\177]
= 92.3 + 80.0 = 172.3kPa.
S,.
= 0.25x10.0x
Iog(172.3/92.3)/(1 + 0,8) = 0.377m.)))
240 Soil Mechanics: Basic Concepts and Engineering Applications

Time (month)
0 1 0 20 3O 4O
17.5
0.1
0.2
0,3
0.4
U 0.5
0.6
0.7
0.8
0.9
1

Figure 6.18. Example 6.9.

From Equations 6.39 and 6.42 De, n and F(n) are calculated:
6.35,
De = 3.0\303\227
I. 128 = 3,384 m, n = D /D w = 3.384/0.3
e
= 11.28, F(n) = ln(\177. 1.28) - 3/4 = 1,673.
6.20: T/\177= 3.20 \303\227
0.5 ! 3.384
2 = 0.1397and from
Applying Equation Equation 6.38:
Uh = 1 - exp(-8 \303\227
0.1397 / 1.673) = 0.4873 = 48.73%.
= = 0.02575 = --\177U --- 0.1811 = 18.11%.
\177tUt,2/4
T 5.15\303\2270.5/10.02
v v

With no contribution from horizontal flow the degree of consolidation after 6 months is
I8.1 i%. The overall degree of consolidation is calculated from Equation 6.48:
U = 1- (1- 0.4873)(i- 0.1811) = 0.5801 = 58%,Sc(6 months)= 0.377 m \303\2270.580 l = 0.219 m.
The results are tabulated below and shown in Figure 6.18, from which tg0 = 17.5months.

Time (months) U\177 Ua U Time (months) U\177. U,\177 U

l 0,0730 0.1054 0.1707 10 0.2337 0.6716 0.7483

2 0.1045 0.1996 0.2832 11 0.2452 0.7062 0.7782
3 0.1280 0\1772840 0.3756 12 0.2561 0.7372 0.8045
4 0.1478 0,3594 0.4541 17 0.3048 0.8494 0,8953
5 0.1653 0.4269 0.5216 17.5 0.3092 0.8575 0.9016
6 0.[811 0.4873 0.5801 18 0.3136 0.8653 0.9075
7 0.1956 0\1775414 0.631I 24 0.3621 0.9309 0.9559
8 0.2091 0,5897 0.6755 30 0.4049 0.9646 0.9789
9 0,2218 0.6329 0.7143 36 0.4435 0\1779818 0.9899

6.3,6 Numerical solution using the finite difference method

The finite difference approach be a convenient
has been proved to
method to model the con-
soIidation theory with reasonableet
accuracy (Olson al,, 1974). O Ison (1.998) reported that the
incremental small strain approach seemed to work acceptably well and to execute faster than

the more complicated solutions, The shnplified version is based on the finite difference appro-)))
 One Dimensional Consolidation 241

(a) (b)

Figure Finite
6.19\303\267 difference grid for calculation of us: (a) vertical drainage. (b) radial drainage.

ximation of Equations 6.16 and 6.36

using the depth-time grid shown in Figure 6.19. For ver-
tical drainage the clay layer is divided into equal parts of thickness of Az. The time interval at
which the excess pore pressure is required is \177. From Taylor's theorem:
au e 1 -
(Ui+i'J l\177i'j
) (6.49)
\177)t \177t

--
(Ui'J +l +Mi.J -I --2Ui,j) (6.50)
0Z2 (\177)2

Substituting Equations 6.49 and 6.50 into Equation 6.16:

Ui+l,j
.... Ui,j
+
c v At
(tCt'J +I
+
ui,j_ [ -2u i , j) (6.51)
(\1772\177)2

As higher order derivatives bare been neglected care must be taken in the selection of/\177t and
2
3z so that the term c,,At/(Az) is approximately 1 / 6 but certainly not 1 / 2 (Craig,
exceeding
1997). This equation applies only within the grid; for the points on the impermeable boundary:

--=O,or -ui,j+l )= 0 at\177d = tg,

ui..\177._ \177 i,j+ i and hence,
az -\177(\177\302\242i,j-1

+ \177
c\177t
\342\200\242
ui+l. j =ui, j (2u',J +I -2ui.j) (6.52)
(\177g,)2

For a layered soil with different per.neabilities and consolidation characteristics, Scott(1963)
showed that the first and second derivations are given by:

k Ou
e I kr k\177 1
(6.53)
o,

1 kT + kB 2k r
,1;
a2ue - \177
2k,\177
\342\200\242
(6.54))))
2 (\177'-7\177,2)(\177Ui
2 (\177.)
\"j+l u;'J-I -2u,.,j)
Oz kT+kf\177 +kg+kg
242 Soil Mechanics: Basic Concepts and Engineering Applications

where the symbols T and B represent the top and the bottom layers respectively. Substituting
these into the primary differential equation of consolidation (Equation 6.16) we obtain:
k
T
+k B \177Xt 2k T
2kB
Ui+l,J =ui,j
\177
+1 + ui ,j - 1 -2ui,j)
kT +kB(cvT/CvB) (\177)2 (kT
+k-\177--\177u\177.) kT +k B

(6.55)
For radial drainage the typical radius-time grid of Figure 6.19(b) is used to obtain:

Ch\177 \177 Ar
+ (6.56)
+\177r
(ui, l+l -ui,j

Example 6.10
A saturated clay layer of 7.5 m is subjected to the vertical loading distribution tabulated below.

Depth (m) 0.00 1.5 3.00 4,5 6.00 7.5

A\177 (kPa) 200.00 145\17745 114.28 94.11 80.00 69.56

The clay layer is resting o,1 an impermeable rock, but the upper boundary has free draining
conditions. Using the finite difference method obtain the values of excess pore water pressures
at the end of each month during one year of consolidation and calculate the average degree of
consolidationafter one year. Az
= 1.5 m, cv = 2.6 m2/year,
Solution:
It is reasonable to assume that the applied load represents the initial distribution of excess pore
pressurethroughout the depth of the soil except at the upper boundary where it is always zero.
2 = 2.6x (1/12)/1.52= 0.09629.Therefore, the points the
cvAt/(Az) for within grid:
= + + ui, -
2ui. j ).
ui+t. j u\177.j O.09629(ui.j+ 1 j_ 1

For the points on the impermeable boundary at the depth of 7.5 m:

- -
ui+l, j u,, ) + 0.09629(2ui,
j+ I 2ui. j )-
The excesspore pressures (in kPa) are shown in the table below and plotted i,t Figure 6.20. It
is evident that the excess pore pressure in the vicinity of the closed boundary slightly increases
over time, but eventually witl decease.
Sample calculation for depth
= 4.5 m (/= 3), and t = 8 months (i = 8):

ue
= 95.54
(corresponding to i = 7) + 0.09629 (10!.23 + 84.35- 2 \303\227
95.54) = 95.0 kPa,
Calculation of average degree of consolidation:
The average initial excess pore pressure of the layer:
ut average = Area of isochrone/ 7.5=
(0.00 / 2 + 145,45 + I i4.28 + 94.11 + 80.00 + 69.56 / 2)1.5] 7.5= 93.7kPa.
The average excess pore pressure of the Iayer at the end of 1 year is:
ire.average
= A\177-ea of isochrone / 7.5 =
(0.00 / 2 + 56.67 + 86.98 + 91.75 + 86.48 + 83.39/ 2) 1.5/ 7.5=72.7 kPa.)))
 One Dimensional Consolidation 243

Excess pore pressure u e (kPa)

200 150 100 50 0

0
1

t = t month\177'\177 \177\"
t = 12 month

5 \302\2433

Figure 6.20. Example 6.10.

_. 0 1 2 3 4 5 6

0 0.00 0.00 0.00 0.00 0.00 0.00 0.00

1 145.45 128,44 114.81 103.73 94.61 87.00 80.58
2 114.28 115.34 114.61 112.77 110.26 107\17738 104.33
3 94.11 94.69 95.30 95.77 96.04 96.08 95,91
4 80.00 80.35 80,88 81.53 82.25 82.98 83.69
5 69.56 71.57 73.26 74,73 76.04 77.23 78.34
-.
J+
7 8 9 10 11 12

0 0.00 0.00 0.00 0.00 0.00 0.00

1 75.11 70.39 66.29 62.69 59.51 56,68
2 101.23 98. f7 95.19 92.32 89.59 86.99
3 95.54 95.01 94,34 93.57 92.71 91.79
4 84.35 84.95 85.47 85.91 86.27 86.55
5 79.37 80.33 81.22 82.04 82.78 83.45

Average degree of consolidationU = 1 - 72.7 93.7 / = 0.224 = 22.4%.

6.4 APPLICATION OF PARABOLIC ISOCHRONES

6.4.1 Umfonn initial excess pore pressure in open and half closed layers
The general shape of the isoch,ot\177e shown in Figure 6.8 may be approximated by a parabolic
equation. This leads to a,\177other U-Tv relationship that is rather different from the curve (1)
shown in Figure 6.9. Figure 6.21(a) shows two isochrones in the ue-z coordinate system. The
middle portion of the isochrone corresponding to a low value of Tv\177is approximately a vertical)))
24-4 Soil Mechanics: Basic Concepts and Engineering Applications

II
U\177 e

oI: <
_x d
Tv -I-
c
_\177 \177//\177 /Tv>Tc

\177u \177
i ue. d= AO' U
e.\177l
\177k
I

I
Z I Z

(a) (0)

Figure 6.21. (a) lsochrones for low and high values of T\177.,(b) parabolic isochrones.

line, indicating high excess pore pressures and that the process of consolidation has not yet
progressed into this part. Identical parabolas may be used to approximate the Iower and upper

portions of this isochrone. For a specificrange of Tv the parabolas are tangent to the vertical
line drawn from the zero consolidation point. The maximum depth of the tangent point (Zo) is
equal to half the thickness d of the layer. If Tc is assumed to be the time factor corresponding
to zo = d, the isochrone with a time factor of Tv2 > Tc represents a more deveIopedcaseand is
well approximated by a parabola. Assuming a paraboIic equation of the form of ue = az2+ bz +
c for T,, T,.,
< then =
at z 0, ue = =
0, and at z zo (unkrtown}, ue = ui. Substituting these condi-
tions in the equation of the parabola and noting that the latter point is the maximum, we have:

ue
=-\177
ui \1772 2ui
+__\177 \177
(6.57)

From the properties of a parabola the volume change is expressed by:

3,V- S,.x I unit area - my x area outside of the isochrone = tnv x Zo x ui / 3
The basicconcept of co,\177solidation states that the rates of volume change and flow are equal
Oh k u
ui
dz\302\260 i 2 2ui
mv -l\303\227k\303\227i=k --=\177 \177
(-_-Tz +--z)/ z, atz=0:
dt 3 \177z y\177 zo zo

dz
o ut -
2ku 6k
mv
i
, thus: z n-
dz\177,
=
dt 3 Zoy w dt

Integrating the equation above between t = 0, z = 0 and t = t, z = zo:

= 6,20 for
z o \177[12kt/mv\302\245 w , using Equation
= (6.58)
% \177
Replacing z\177in Equation 6.57:
- z2 2z
ue = u
i (-- + \177) (6.59)
12Cvt)))
 One Dimensional Consolidation 245

Equation 6.59 may be represented in U-T,, coordinate system using the main definition for the
average degree of consolidation:
U = = --\177
(zou i/3)/(d\303\227u i) \177/3d 2\177/3 (6.60)

This equation is valid until zo becomes equal to d. At this depth:

= = d, or z = 2 =
d, thus = [ [ 12 and t = d 2 / 12cv
Zo
l\177-cvt o l\177vd Tv

For the values of T\177,greater than 1/ 12 the equation of parabola is:

2Ue-d
ue-d z2 +\177Z
Ue
=
2 d
(6.61)
d

where Ue-d is the excess pore pressure at the centre of the layer as shown in Figure 6.21(b),
Equating the rate of change of volume with d\177e flow rate a\177d integrating:

1 3Cvt
u e_\177t = u i
exp(\177---\177)

Replacing Ue\177d in the equation of the parabola (Equation 6.61):

2 1
= u
z 2z 3Cvt
u + (6.62)
e i
(--\177_ --\177-)exp(\177 --\177-.)

In the U-T\177 coordinate system:

U = 1- 2exp(1 - 3T v ) (6.63)
3 4
Equations 6.60 (for Tv < 1 / 12) and 6.63 (t\177r Tv > 1 / 12) are shown in Figure 6.22. If the lin-

ear portion of the curve is extendeddownwards it will intersect the U = 1 axis at Tv = 0.75:

U=0%

N T
v
= 0.75

,,

Figure 6.22. Reference point to calculate c\177,,)))

246 Soil Mechanics: Basic Concepts and Engineering Applications

This could be a reference point to calculate c,, from the square root time-settlement curve pro-
ducedfrom experimental results. If the time corresponding to T,,= 0.75is \177tthen:
2
c v = 3d / 4t l (6.64)

It may be shown that the average de\177ee of consolidation estimated by parabolic isochrones is
slightly higher than thetraditional method, and results in a higher value of effectivestress
(higher settlement) and a lower value of excesspore pressure.

Example 6.11

Resolve Example 6.6 using the parabolic isochrones.

Solution:
Calculations are carried out using Equation 6.62 for (only for z / d from 0 to
Tv
-- 0.35 1.0)and
are tabulated below, The results are plotted for thickness in Figure 6.23.
the whole
Average U from Equation 6.63: U =1-(2/3)exp(l/4-3\303\2270.35) = 0.700 = 70.0%.

/ d -
z u:/ u\177 U\177= l u\177/ ui U.z%(parabola) Uz%(Exact)
0.0t3 0.0000 1.00013 !00 1 O0
0.25 0.1966 0.8034 80,34 79.43
0.50 0.3370 0.6630 66.30 62.02
0.75 0.4212 0.5788 57.88 50.40
1.00 0.4493 0.5507 55.07 46.33

Example 6.12
\"9
Using the following consolidation test data determine cv from the root time plot in re-/year.

0 20 40 60 80 100
0

0.25
0.5
0.75
1
1.25
1.5
1.75
2

Fibre 6.23.Exan'tple 6. I 1.)))

 One Dimensional Consolidation 247

Root of time(m in)

0 1 2 3 4 5 6 7 8 9 10 11
O\177

0.2
0.:3
0.4
0.5 --
0.6 \177
0.7
08
0.9

F\177gure 6.24. Example 6.12.

Ttme (rain) Total ZX/-/(ram) Tirae (ram) Total AH (ram) Tirae (rain) Total 4,/t (tara)

0.25 O. 138 6,25 0.524 49 0.894

0.50 O. [ 96 9 0.597 64 0.900
l.O0 0.277 16 0.775 81 0.900
2\17725 0.375 25 0,855 100 0.900
4.00 0.4[ 5 36 0.890

Solution:
The given are plotted in Figure
data 6.24. from which: tt = (3.25)- = 10.56 rain,

2d (average) = [20.0 + (20.0 -0,9)] / 2 =19.55. d = 9.77

,6
ram.
=
c,, [0.75(9\17777)2 / 10,56] x x
60 24 365
x x 10 = 3,56 mZ/year,

6.4.2 Linear initial excess pore pressure in a half closed layer

The solution of consolidation

one-dimensional using parabolic isochroneswith linear initial
excess pore pressure is limited case where the maximum
to the initial excess pore pressure is
appliedat the closed boundary while at the free draining boundary its value is zero. The re-
verseconditions cannot be approximated by parabolic isochrones due to the increase of excess
pore pressure on the closed bounda\177 in low values of Tv. If the excess pore pressure on the
free draining boundary is not zero and has a value less than the excess pore pressure on the
closed boundary then a superposition method can be used.The initial linear excess pore pres-
sure can be divided into two loading of uniform and linear. The magnitude of the uniform
pressure is equal to its minimum value on the free draining boundary. Its linear part starts from
zero at the free draining boundary and linearly increases to its maximum at the closed bound-
ary. The magnitude of this maximum is equal to the difference between boundary values. Ba-
sic assumptions for the parabolic isochrones are shown in Figure 6.25. For Tv values smaller
than a specific value T,. each isochroneis tangent to the initial excess pore pressureline (point
7) and perpendicular to the closed boundary at B. For a value of Tv > T,. the isochrone is)))
248 Soil Mechanics: Basic Concepts and Engineering Applications

II e ,\177,

A BC D E

Figure 6.25. Approximation of isochrone with parabola for linear initial excess pore pressure.

located under the initial excess pore pressure line and is perpendicular to the closed boundary
2
at point E. To establishthe isochrone for Tv = Tc we assumeue = az + bz + c. At z = 0, ue = O,
and the slope of the curve is ui/d. Hence, the equation of parabola is:

ui 2 Ui
z + z (6.65)
2d 2 d
At z = d, Ue-d = ui / 2 therefore: Sc =m v
u i 12-
(d \303\227 2/3d \303\227
ue_ d )
=
duim v [ 6. Equating Sc to
the integration of the rate of flow:

ku i d2 1
duim v
_tf t = kui t
-'>t=t,
6 \177Oy\177--\177d Yw d :6c---\177-or T\177,.=\177-

For the values of T\177< 1 / 6 the typical equation of a parabolic isochrone is constructedbased
on the geometry shown in Figure 6.25 using Zo as the depth for the point T:

- 2
zo ) +
ue (z -zo) (6.66)
+-\177(z-
2d(d-Ui-zo) uiz\302\260d

In order to relate Zo to time we first find the excess pore pressure at the closed boundary. Re-
placing z = d in Equation 6.66:

= FB=
Ue_ d ui(d + zo)/2d (6.67)
particular time is Sc= my(area
The settlement at this of TAC - area of TBC).Using the geome-
try of the figure we obtain:

-
Sc - mvui(d z\302\260)2
(6.68)
6d

Using a unit area of soil, then volume change and settlement are numerically equal and the rate
of volume change is equal to the rate of flow. In other words, settlement is equal to the
integration of the rate of flow:
_ ku ku i
Sc =
mvUi(d Zo)2 = t\177@t
i = --t ' this yields
\) a relationship between zo and time:
6d Y Y wd
\1770 wa\"
 One Dimensional Consolidation 249

(d - z 0)2
=
6Cvt (6.69)
Combining Equation 6.68 and 6.69 we get:

mvui(d-
Sc =
- mvui
(6.70)
z\302\260)26d \303\2276cvt6d=-mvUiCVtd

The final settlement of the layer is: Sc(final) =

mvuid / 2 ; from the basic definition of average

degree of consolidation: U = So(at specific time) / final settlement, thus:

U =
2 = 2T for 1 / 6
2cvt/d v T\177.< (6.71)
Using a similar procedure for the values of Tv > 1 / 6 the following equations are obtained.

= -ue-a z 2 +
2Ue-a
z
u
e (6.72)
d-\177 d
u 1
= i -
Ue-a 3T\177,) (6.73)
\177-_exp(-\177

d
2Ue-d d
Sc ) at any specified time (6.74)
=mv(.u-\177 3

U = t-2exp(1-3T\177) (6.75)
3 2
Equations 6.60 and 6.63 are represented by curve PI in Figure 6.26 that approximates the ex-
act solution shown by curve (1) in Figure 6.9. Similarly, Equations 6.71 and 6.75 are repre-
sented by curve .02 which approximates the exact solution shown by curve (2) in Figure 6.9.

0 0.1 0.20.30.40.50.60.7 0.8 0.9 1 1.1 1.21.31.41.5

(1.2
0.:3

'
U 0.5
0.6
0.7

Figure 6.26. Relalionship between U and Tv using parabolic isochrones.)))

250 Soil Mechanics: Basic Cottcepts and Engineerb\177g Applications

Example 6.13

A slurry material comprised is deposited in a large tagoon with an imper-

of fly ash and water
meable surface. The material thickness deposition is 4 m. If the density
afterof the slurry is
1.25 M\177m 3. my = 7.5 m2/MN, and cl, = 5.5 m /day, determine: (a) the ttme required for 90%
consolidation, the corresponding settlement and the excess porepressureat the surface of the
impermeable layer, (b) the equations and plots of isochronesafter 8 and 16 hours, the average
degrees of consolidation and the reIevant surface settlements, (c) the equation of excess pore

pressure on the surface of the impermeable boundary in terms of time and a plot of the results.

Solution:
(a) initial excess pore pressure is linear
The and varies from zero at the surface of the slurry to

ui atthe surface of the lagoon: u i = ()' -yw)\303\2274.0 = (1.25

-1.0)x9.81\303\2274.0
= 9.8 kPa.
For U 90% the value of T\177.is calculated
= from Equation 6.75:
U = t
- (2 / 3) / 2 - 3T ) = 0.9 \177 = 0.80,
exp(l t. Tx,
2 = = 0.8 --\177t = 2.3 days.
Tv =
c\177,t/d 5.5xt/16.0
Sc (final) =mvu\177d/2
= 7.5x 10 -3 xg.Sx4.0/2 = 0.147 m, Sc(90%)
= 0.147x0.9 = 0.132 m.
From Equation
- 3 x 0.80) = 0.7 kPa.
6.73: ue_ d = (9.8/ 2)exp(0.5

(b) T\177.= (5.5 x 8.0/24)/16.0 = 0.!i46< 1/6 Calculate Zo from 6.69:

Equation
2
(4.0 zo) =6cvt=6\303\2275.5xS/24--\1774.0-z o=3.31m,zo-O.69m.
Substitute z,\177in Equation 6.66 to find the excess pore pressure parabola'.
u = -0.37z :\177+ 2.96z -0.176. Similarly the equation of isochrone at t= i6 hours is:
e
u e = -0.254z
2 + 2.031z.Note that in this case - 0.2292 > 1 / 6. The results are ia
Tv plotted
Figure 6.27(a).

Ue (kPa)

10 9 8 7 6 5 4 3 2 1 0

0 10
0.5 9
1
8
7

2 \177 5
2.5 \177'\177
=
4
3
3
2
3.5 1
4 0
(at
10 2O 3O 40 50 6O)))

t (hour)

(b)

Fi\177mare 6.27. Example 6.13.

 One Dimensional Consolidation 251

For t = 8 hours: Tv = 0.1146 < / 6. From Equation 6.71:

l

U = 2 T\177= 2 \303\227
0.1146 = 0.2292 = 22.92%, Sc= 0.2292 O. 147 = 0.03369 m = 33.7 ram.
\303\227

For t =16 hours: Tv = 0.2292 > 1 / 6, From Equation 6.75:

U -- 1- 2/3\303\227exp(l / 2 - 3\303\227t3.2292) = 0.4474 = 44.74% ,S = 0.4474\303\22713.i47
c \303\2271000= 65.Srnm.

5.5 m2/day = 5.5 / 24 = 0.2292m2/hour.

(c) c,,=
2 = (1/ 6)\303\227 = 11.63 hours. Substitute
For T\177
< 1/ 6 or t < T d /
v cv 16.13/0.2292 Zo from Equation

6.69 into Equation 6.67 to obtain:

u = (u, / 2d)(2d - - (9.8/8.0)(8.0-\177/6xO.2292t = (9,8/8.0)(8.0-t.t7347).
e_ d 6\177-\177vt)

For G, > 1 / 6 or t > 11.63 hours use Equation 6.73, but substitute Tv in terms oft:

Ue_ d
= (9.S/2)exp(0.5- 3x0.2292t/16.0) = (9.S/2)exp(0.5- 0.043t).
The results are plotted in Figure 6.27(b).

6.5 LIMITATIONS OF ONE DIMENSIONAL

CONSOLIDATION THEORY

6.5.1 Correction of primary consolidation settlement

In the derivation of the U-T,, relationships it was assumed that the increase in the external load
is instantaneous. However, construction work is normally a gradual process and may be ideal-
ized as shown in Figure 6.28, where tc is the time when stress reaches its
the applied vertical
full amount An. When using a numerical method, time dependent load
any type of non-linear
can generally be incorporated in the solution. Terzaghi's method for the correction of the time-
settlement relationship (curve (b), Figure 6.28) for an instantaneous loading is based on the
The corrected settlement t > t,. is to on the un-
following assumptions. at equal the settlement
corrected curve correspondingto t/2. For t < tc further correction is needed by considering the
load ratio at the time of interest. The load ratio is the ratio of the Ioad at the time t to the final
load Ac\177. The settlement corresponding to t/2 is multiplied by this ratio. The curve (c) in

\177'k\177/t o
/ 2 Lo\177ad tc

(b)\177

Figure 6.28. Terzaghi's correction lbr 6me dependent loading.)))

252 Soil Mechanics: Basic Conceptsand Engineering Applications

Figure 6.28 is the corrected time-settlement relationship. This method implies that for a speci-
fied U the amount of time is twice of the time required for instantaneous loading.

Example 6.14
At the centre of a 2
clay layer the effective vertical
m thick saturated stress is increased from
60 kPa to 160 kPa of 6 months. Plot the variation
within a period of consolidation settlement
against time up to 48 months. The layer is half closed and has the following properties:
Cc = 0.15, cv= 1.5m2/year, e0
= 0.7.

Solution:
Sc(final) = 0.15 \303\227
2000\303\227log(160.0 / 60.0)/(1 + 0.7) = 75.2mm.
Results computations are tabulated
of the below and plotted in Figure 6.29. Sample calcula-
tions for and 12 months
3, 6, are as follows:
2 =
For t =3 months, Tv =cvt/d =0.0937 =\177U 2/4-\177U
1.5\303\227(3.0/12.0)/2.02 =0,345,
=
Sc 0.345 \303\227 75.2 = 25.9 mm. Corrected Tv, U and So:
=
1.5\303\227(1.5/12.0)/2.02 = 0.0469 = \177U 2/4 --\177U = 0.244,
T\177.

U (corrected) = 0.244(3.0 / 6,0)= 0.122, Sc (corrected) = 0.122 \303\22775.2 = 9.2 mm.

For t = 6 months, Tv = 1.5 \303\227 (6.0/12.0)/2.0'2
= 0.1875 = 7\177U2 /4 -\177 U = 0.488,

Sc = 0.488\303\227 75.2 = 36.7 roan. Corrected T,,, U and So:

For 6.0 / 2 = 3.0 months we have already calculated Tv and U: Tv = 0.0937, U = 0.345,

U (corrected) =
0.345(6.0 / 6.0 ) = 0.345,Sc (corrected) = 0.345 \303\227
75.2 = 25.9 ram. For t = 12
months, T =0.375 = -0.933tog(I-U) -0.085 --\177U = 0.679,
v =1.5\303\227(12,0/12.0)/2,02

Sc = 0.679\303\227 75.2 = 51.1 mm. Corrected Tv, U and St:

For 12.0/ 2 = 6.0months we have already calculated Tv and U: Tv --O.1875, U = 0.488, no fur-
ther correction required, thus Sc= 0.488\303\227 75.2 = 36.7 mm.

Time U U (nun)
S\177- Sc (ram)
(Month) Uncorrected Corrected Uncorrected Correcled

0.5 0.141 0,008 10.6 0.6

I 0.199 0.023 15,0 1.7
2 0.282 0.066 21.2 5.0
3 0.345 0. 122 25.9 9.2
4 0.399 0.188 30.0 14.1
5 0.446 0.263 33.5 19.8
6 0.488 0.345 36.7 25.9
9 0.598 0,423 45.0 31.8
12 0.679 0.488 51.1 36.7
18 0.798 0.598 60\1770 45.0
24 0. 873 0,679 65.6 51.1
30 0.920 0.745 69.2 56.0
36 0.949 0.798 71.4 60.0
42 0.968 0.839 72.8 63. I

48 0.979 0.873 73.6 65.6)))

 One D#nensionaf Consolidation 253

Time (months)
0 20 40 60
o
lO
2O

k-'\177
\177,, Corr ect ed

03 60 \"\"
Uncorrected \177
70 \177 _
80

Figure 6.29. Example 6.14.

6.5.2 Sources of inaccuracy and three-dimensional consolidation

The main sources of inaccuracy

in the prediction of both the magnitude and rate of consolida-
tion setdement
compared with in-situ measurements are (Duncan, 1993):
1. Inaccuracies in the measurement of consolidation characteristics such as preconsolidation
pressure p'c, compression index Co, coefficientof volume compressibility my, and the coeffi-
cient of consolidation c\177in the laboratory conditions.
2. Difficulties in evaluating the drainage conditions in the field.
3. Shortcomings in the conventional consolidation theory.
Inaccuracies in the magnitude of p'c, even with undisturbed specimens, could occur due to
adoption of the 24 hour-based e-logo\" as being representative of field behaviour. Secondary
consolidationmay alter the overconsolidation ratio from 0 to a large value under the same ef-
fective vertical stress as suggested by Bjerrum (1967), introducing time-related consolidation
curves. Point A in Figure 6.30 represents the state of stress on the virgin compression line con-
structed in laboratory conditions. As time passesbeyond the 24 hour mark the state of stress
drops vertically to point B on a consoIidationcurve t, indicating a secondary consolidation set-
tlement. A point in the field with the same state of stress as B will, if it is tested in the labora-

tory, pass an apparent preconsolidation pressure \177'm corresponding to an overconsolidation ra-

tio of \177',\177
/ \177\"0. The overconsolidated soil can also be created by the changes in the ground
water level, as discussed earIierin this chapter.
The finaI primary consolidation is calculated either using Cc or my. Accuracy of the predic-
tion depends on the selection of Cc and my as well as the method of obtaining the incremental
effective vertical stresses due to the surface loading. The parameter involved in the calculation
of degree of consolidation(rate of settlement) is cv, which is determi\177ed in the laboratory. In-

accuracies arise from both the method used in the laboratory as well as the evaluation of the)))
254 Soil Mechanics: Basic Concepts and Engineering Applications

t=0

', \\ \177\\\\.\\

- \177 log\177
d'0 d 'm

Fibre 6.30. The concept of app\177ent overconsolidation (Bje\177m, \177967).

drainage length. Most of the standard codes both root time and log time methods.
recommend
However, theexperimental results obtained
laboratory do not perfectly satisfy
in the the theo-
retical equations and the cv obtained from the methods above become different within a small

range. In some clays this difference may give rise to a high value, as reported by Duncan

(1993) for San Francisco Bay mud (highly plastic organic clayey silt), where the root ti\177ne
method gave 50% more than the value determined by the log time method. It must be noted
that the magnitude of cv is dependent on the stress level and changes during the dissipation of
the excess pore pressure,considering its definition by Equation 6.17 (c v = k l(m\177.yw) ). As a
result of the progress of consolidation the coefficient of permeability k decreases. Further-
\177nore, the decrease in volume results in an increase in strength and resistance to volume
changeand consequently, my decreases. In general the value of c\177,decreases as the effective
vertical stress approachesthe preconsolidation pressure. However, in normally consolidated
clays, Cv may increase with the increase of effective vertical stress along the virgin compres-
sion line. Sometimes reductions in k and my occur such that cv remains constant during the
consolidation progress.
Computations are simplified if the drainage length (or the average thickness of the speci-
men) is calculated using identical equations for the field and the laboratory. Olson & Ladd
(1979)suggest the use of following relationship for a constant average thickness:

H = Ho ST
(6.76)
2N o

where H0 is the initial thickness, ST is the ultimate settlement and ND is the number of free
draining boundaries.
One of the major basicassumptions of the conventional consolidation theory is the assump-

tion of a linear elasticsoil, which is inherent in the different{al equation of one-dimensional

consolidation. The elasticproperties Es and Nt are replaced by m,, in the form of proportionality)))
 One Dimensional Consolidation 255

of volume change and effective vertical stress. From the principles of elastic theory it can be
shown that, for zero lateral strain, the magnitude of mv is given by:

-
mv -
(1+ la)(1
(6.77)
(1-g)E s
However, soil is not a linear elastic material and the inclusion of non-linear behaviour in the

main differential equation is only possible by means of numerical analysis (Mesri & Rokhsar,
1974). Note that the eqogo-' curve is non-finear, but this is only used to calculate the final set-
tlement and is not incorporated in the time-settlement solution.
Conventional consolidationtheory is applicable for small strains and cannot be applied to
hydraulic fills and other slurry materials where the large strains control the consolidation pro-
gress. For large strain consolidation, most of the methods developedfor the estimation of final
settlements are based on the mathematical treatment of data obtained from the field measure-
ment at the early and advanced stages of consolidation (Tan et al., 1991).
It must be noted that most consolidation settlements in the field have a two or three-

dimensional nature. To better representfield conditions, attempts have been made to correct
settlements predicted using one-dimensional theory. Skempton & Bjerrum (1957) proposed
the following equation for the computation of final primary consolidation settlement:

Sc = \177 (mvAuAz) i (6.78)

where Az is the thickness of a finite layer and Au is the initial average excess porepressure (fi-
nal increment in effective vertical stress) in the layer and is calculated from Equation 4.27 by
assuming B = 1:
Au = AcJ
3 + -A\177
A(A\302\2425
t 3)

where Acq and A\1773 represent the incremental stresses (total) in the vertical and lateral direc-
tions respectively, Substituting Au in Equation 6.78 we obtain the three-dimensional consoli-
dation settlement:
i=tl

S,,(3
- D) = [(AAc
'\177'\177 h + (1- A)A\177 3 )mvAz]i (6.79)
i=I
The combination of Equation 6.79 and 6.13 (one dimensional consolidationsettlement So) may
be written in the following form:

S\177(3
- D) = \177.S,.\177 (6.80)

i=rl
where: L = A + 0:(1- and, \177=
Z
A) (A\177ramvAZ)s
/
i=l

The parameter \177.varies with pore pressure and c\177,the latter being dependent on the vertical and
lateral stress distribution within the consolidated layer. The variation of \177,with the pore pres-
sure coefficient A for both circular and strip footings has been presented by Scott (1963).
Equation 6.80 can be applied to a circular footing with reasonable accuracy assuming that a)))
256 Soil Mechanics: Basic Concepts and Engineering Applications

1.2

0.8

0.6

Circle
Strip
--

o.2

0 0.2 0.4 0.6 0.8 1 1.2

Pore pressure coefficient A

Figure 6.3 I. Settlement coefficients for 3-dimensional consolidation.

uniform load transmitted from the circular footing will create triaxia] stress conditions in the
soil underneath, For a strip footing, Scott (i963) modified Equation 6.80 to the form:

L = N + \177(i- N) (6.81)

where N = (-\177'/2)(A -
1/3) +0.5, The coefficientk varies from 0.25 to 1,2 for a range of pore
pressure coefficientA between 0 and 1.2 within a practical range of thickness of clay layer H
to the width B ratios (Figure 6.31),Balasubramaniam
of footing & Brenner (198 l)challenged
this method and indicated that Equation 6.79 is strictly valid only for triaxial stress conditions,
and that a stress path method (Lambe & Mart, 1979) can better take into account the effects of
lateral deformation on vertical settlement.

6.6 PROBLEMS

6.1 Data obtained from a laboratory consolidation test are tabulated as follows:

\177 (kPa) 20 50 100 200 400 800

Total AH (nm\177) 0,23 0.87 1.90 3,62 5.55 7.25

G\177.
= 2,70, H0 = (initial ttfickness at zero pressure) = 22.5 ram, w = (moisture content at

the beginning of the test) = 0.78.Plot the e-log( curve and calculate Ce.)))
 One Dimensional Consolidation 257

Answer: 0.832
6.2 An open layer of clay 4
m thick is subjected to loading that increases the average effec-
tive vertical 185 kPa to 310 kPa. Determine (a) the total settlement,
stress from (b) the
settlement at the end of one year ' (c) the time in days tbr 50% consolidation(d) the time
, 2 \177'
\177) mm of settlement
m days for _5 to occur, my= 0.000_5 cv = 0.75 m /year.
\177)
m/k_N,

Answers: 125 ram, 6i ram, 383.5 days, 61 days

6.3 Data obtained from a laboratory consolidation test are shown in table below:

Time (rain) 0,25 1 4 9 16 25 36 81 1440

Total AH (ram) 0.622 1.244 2.468 3.400 3.838 3.970 4.000 4.051 4,100

100 kPa,200 kPa, H0 = 23.6ram. Determine:

o-'1 = (a) cv from the root time plot in
o-'\177=
m/year, (b) cv from the log time plot m m/year. (c) k iu m/s.
-l\302\260
Answers: (root time) 4.9 m2/year, (log time) 5.2 m2/year, 2.8 \303\227
10 m/s

6.4 In a one-dimensional consolidation test the time required for 50% consolidation has been
measured at 154 seconds (through the observation and measurement of pore water pres-
sure).The settlement of the sample at the end of the test was 2.5 rnm. o-'0 = 60 kPa, o\"1
= 120 kPa, e0 = 0.65, H0 = 20 ram. Determine (a) the time for 90% consolida-
required
tion, (b) the coefficient of permeability in m/s, (c) the compression index,
-7
Answers: i i. 1 rain, 2.65 10
\303\227 m/s, 0.685
6.5 For a 4 m layer of the clay of Problem 6.4, how long would it take to reach 50% degree
of consolidation under the same drainage, physical and stress conditions? What will be
the settlement of the clay layer at this stage?
Answers: 71.3 days, 250 mm
6.6 A surface 60 kPa is applied on the ground surface over a large area. Thesoil pro-
load of
file consists of a sand
layer 2 m thick, the top of which is the ground surface, overlying a
4 m thick layer of clay. An impermeable boundary is located at the base of the clay
layer. The water table is 1 m below the ground surface. If the preconsolidation pressure
for a sample of soil from the mid-ly3int of the clay layer is 60 k.Pa, calculate the consoli-
dation settlement of the clay layer, The properties of the soil section are: sand: 9d,y = !.6
3 3 3
Mg/m , P.\177at = 1.9 Mg/m , clay: Psat = 1.65 Mg/m , e0 = 1.5,Co= 0.6, C\177.= 0.i.
Answer: 374 rnm
6.7 A soil profile consists of a sand layer 2 m thick, whose top is the ground surface, and a

clay layer 3 m thick withboundary located at its b\177se. The water table is
an impermeable
at the ground surface. A widespread load of 100kPais applied at the ground surface.

Construct isochrones correspondingto 10%, 50%, and 90% consolidation, indicate the
amount of excess pore pressures on the impermeable boundary and determine the
amount of settlement after 2 years. Assume that the soil is in a normally consolidate
state, The propertiesof the soil section are: sand: Ysat = 20 kN/m 3, clay: \"lsat = 16 kN/m 3,)))
258 Soil Mechanics: Basic Concepts and Engineering Applications

eo = 1.3, Cc = 0.5,cv = 6.5

m'/year.
Answers: 100.0 kPa, 77.8 kPa, 15.7 kPa, 0.4!. m
6.8 If the water leveI in Problem 6.7 is Ioweredto the surface of the clay layer; calculate the
consolidation settlement of the clay layer after 6 months if (a) the drawdown is instanta-
neous, (b) the drawdown takes 2 months. Assume there is no surface load and take the
3
unit weight of the sand Iayer 17.5 kN/m after the draw down has taken place.

Answers: 76 ram, 54 mm
6.9 A stratum of clay is 5 m thick and is overlain by 3 m of sand, the top of which is the
ground surface that is subjected to a widespreadIoad of 200 kPa. The water table is 1.5
m below the ground surface, and the pore pressureat the impermeable boundary was
measured to be 242.5\177kPa after 18 months. If the settlement of the ground surface was
230 ram, determine the field values of cv and Co, the final consolidation settIement, and
the settlement and pore pressure at the base of the cIay layer after 3 years, using the con-

cept o\177fparabolic isochrones. The properties of the soil section are: sand: 9d\177. = 1.75
Mg/m\", 9sa\177 = 2
Mg/m 3, clay: Osat = [.95 Mg/m 3, e0 = 0.8.
Answers: cv
= 2.0
re-/year, Cc = 0.3327,570 man, 333 mm, i46.8 kPa
6.i0 In the soil profile of Problem 6.9, vertical drains of diameter 0.3 m are constructedin a

square pattern. It is required that 95% of the combined consolidationbe achieved after
1.5 years. Calculate the required distance between the vertical drains. Ch = 4 m'/year.
Answer: 3.0 m
6.11 In Problem 6.7, calculate the consolidation settlements at 3 months, 6 months and 2
years if the load is increased linearly to 100 kPa over 6 months.
Answers: 70 ram, 200 ram, 360 mm

6.7 REFERENCES
A1-Khafaji, A.W. & Andersland, O.B. 1992,Equations for compression index approximations. Journal
GED, ASCE 118(I): 148-153.
ASTM D-2435. I996. Standard test method for one-dimensional consolidation properties of soils.PA,
West Conshohocken: American Society for Testing and Materials.
Australian Standard AS 1289.6.6.1. 1998, Methods of testing for engineering purposes, method 6.6.t:
soil strength and consolMation tests-dewrmination of one-dimet\177sional consolidation properties of a
soil-standard method. Australia, NSW: Standard Association of Australia\303\267
Azzouz. A.S.. Krizek, R.J. & Corotis.R.B.1976. Regression analysis of soil compressibility. Soils and
foundations 16(2): I9-29.
Balasubramaniam, A,S. & Brenner, R.P. 1981. Consolidation and settlement of soft clay. In E.W. Brand
& R.P. Brenner (eds), Soft' clay engineering: Developmems in geotechnical engineerb\177g 20. New
York: Elsevier Scientific Publishing Company.
Ban-on, R.A. 1948. Consolidation of fine-grained soil by drain wells. Transactions, ASCE 113:718-742.
Bergado, D.T., Anderson, L.R., Miura, A.S. & Ba[asubramaniam, A.S. 1996. Soft' ground improvement
in lowland and o\177her environments. USA: ASCE Press.)))
 One Dimensional Consolidation 259

Bjerrum, 1967. Engineering geology of Norwegian

L. normally consolidated marine days as related to
of buildings. Geotechnique ( 17):8 I- 118.
settlements
British Standard 1377-6. 1990.Methods of test for soils for civil engineering purposes, Consolidation
and permeabili\177, tests in hydraulic cells and with pore pressure measurement. London: British Stan-
dards [nstkution.
Carillo. N. t942. Simple two-and three-dimensionalcasesin the theory of consolidation of soils. Jottrnal
of mathematics and physics 21( I ): 1-5.
Casagrande, A. 1936.The determination of preconsolidation load and its practical significance, discus-
sion 34: Proc, ]st intern, conf SMFE 3: 60-64. Cambridge.
th
Craig. \177997. Soil mechanics.
R\303\267F. 6 edition. London: E & FN SPON.
Duncan, J.M. 1993. Limitations of convemional analysis of consolidation settlement. Journal GE. ASCE
119(9): 1333-1359.
Hansbo. S. 1979. Consolidation of clay by hand shaped prefabricated drains. Ground engineering 12(5):
16-25.
\177
Hansbo. S, 1981.Consolidation of fine-grained soils by prefabricated drains. Proc, 10 intent, co\177
SMFE 3: 12-22. Stockholm.
Hansbo. S. 1987. Design aspects of consolidation of vertical drains and lime column installation. Proc.
\177
9 So\177aheast Asian geotechnical con. 2:8-12. Bangkok, Thailand.

Kjellman, W. 1952, Consolidation of clay by means of atmosphere pressure. Proc. soil stabilization
conj,: 258-263,Mass..
Koppula, S.D. t986. Discussion: consolidation parameters derived from index tests. Geotechnique 36(2):
68-73.
Lambe, T.W. &Marr, W.A. 1979. Stress path method. Journal GE, ASCE 105(GT6):727-738,
Landau, R.E, 1966. Method of installation as a factor in sand drain stabilization design, Highway re-
search board HRR(133): 75-96.
Leonards, G.A. t976. Estimating consolidation settlements of shallow foundations on overconsolidated

clays. Specialreport 163, transportation research board: 13-16.

Mesh. G. 1986. Discussion: postconstmction settlement of an expressway built on peal by precompres-
sion, Ca\177kadian geotechnicaljountal (23)3: 403-407.
Mesri. G. & Castro, A. 1987. The Ca/C\177. concept and ko during secondary compression. Jounzal GE,
ASCE 112(3): 230-2a7.
Mesri, G., Feng, T.W. & Benak. J.M. 1990.Postdensification penetration resistance of clean sands.
Journal GED, ASCE 116(GT7): I095-I 1 t5.
Mesri. G. & Godlewski, P.M, 1977. Time and stress compressibility interrelationship. Journal GE, ASCE
103(5):417-430.
Mesri. G. & Rokhsar, A. 1974.Theory of consolidation for clays..Journal GED, ASCE t00(GT8): 889-
904.
Mesri, G.. Rokhsar, A. & Bohor, B.F, 1975.Composition and compressibility of typical samples of Mex-
ico City clay, (25): 527-554,
Geotechnique
Nagaraj. T.S, & Srinivasa Murth),, B.R. 1985.Prediction of the preconsolidation pressure and recom-

pression index of soils. Geotechnical testing journal, AS1-M (4):199-202,

Olson. R.E\303\267& Ladd, C.C. 1979. One-dimensional consolidation problems. Journal GED. ASCE
105(GTI):11-30.
Olson, R.E., Daniel, D.E. & Liu. T.K. 1974.Finite difference analysis for sand drain problems, Analysis
and desig\177t bz geotechnical engineering, ASCE 1: 85-110.
Olson.R.E. 1998. Settlement of embankments on soft clays. Jounzal GGE, ASCE 124(8):659-669.)))
260 Soil Mechanics: Basic Concepts and Engineering Applications

Parry, R.H,G, & Wroth, C.P. [981. Shear stress-strain properties of soft clay. In E,W. Brand & R.P.
Brenner (eds),SoJ\177 clay engineering: Developments in geotechnicat engineering 20. Ne',u York: E|-
sevier Scientific Publishing Company.
Parsons, J.D. I959. Foundation installation requiring recharging of ground water. Journal of construction
division, ASCE 85(CO2):1-2[.
Pilot, G. 198I. Methods of improving the engineering properties of soft clay. In E.W. Brand & R.P.
Brenner (eds), Soft clay engineering: Developments in geotechnical engineering 20. New York: El-
sevier Scientific Publishing Company.
Powers, ,I.D. 1985. Dewatering-a'\177oiding its unwanted side effects. Underground technology research
council, ASCE. New York.
Raymond, G.P, & Wahls, H.E. 1976,Estimating [-dimensional consolidation, including secondary com-
pression of clay loaded from overconsolidated to normally consolidated state. Special report I63.

trat,sportation research board: 17-23,

Rendon-Herrero, O. 1980. Universal compression index equation. Journal GED, ASCE ]06(GTII):
[ t79-1200.
Rendon-Herrero, O. 1983. Closure: universal compression index equation, Journal GED, ASCE
109(GT5): 755-761.
Rixner, J.J., Kraemer, S.R. A.D, t986. Prefabricated
& Smith, vertical drains. Engineering guidelines,
federal highway administration / 168).WashingtonDC.
l(FHWA-RD-86
Schmertmann, J.H. 1953. Estimating the true consolidation behavior of clay from laboratory test results.
Journal SMFE, ASCE 79 (31[): 26.
Scott, R.E. 1963. Principles of soil mechanics. Reading, Massachusetts: Addison-Wesley.
Skempton, A.W. 19\177,4. Notes on the compressibility of clays. Quart. journal of geological society
(]00):t19-135.London.
Skempton, A.W. & Bjerrum,L. 1957.A contribution to the settlement analysis of foundations on clay.
Geotechnique (7): 168-178.
Tan, T., Inoue,T. & Lee, S. [99t. Hyperbolic method for consolidation \177na]ysis. Journal GED. ASCE
ll7(I1): t723-1737.
Terzaghi, K. & Peck\177 R. B. 1967. Soil mechanics in engineering practice. New York: Jolmra Wiley.
Ter'zaghi, K., Peck, R. B., & Mesri, G. 1996, Soil mechanics in engineering practice. 3 edition. New
York: John Wiley & Sons.
Zeevaert, L. 1957.Foundation design and behaviour of tower Latino Americana in Mexico City. Geo-
technique 7(3): 115-133.)))
CHAPTER7

Application of Limit Analysis to StabilityProblems

in

Soil Mechanics

7.1 INTRODUCTION

The main objective of this chapter is to investigate the stability of a soil structure using the
lower and upper bound theorems of plasticity. Tl\177ese theorems are used to predict collapse
loads where analytical solutions either do not exist or are inconsistent with the governing

equations of mechanics (Bottero al., 1980). They are also usedwhen the deformations
et of the
soil structure are negligible. A lower bound solution provides a safe limit load, whereas an up-
per bound solution estimates an unsafe limit load under which the failure of materia[ has taken
place already. In a lower bound solution only the equilibrium and yield criterion are satisfied,
whilst in an upper bound solution, only the compatibility and the yield criterion are consid-
ered. These solutions, obtained either manually or numerically, bracket the exact solution
witlfin usually acceptable accuracy (Aysen & Sloan, 1991a). These bounds also provide useful
guidance on the accuracy of finite element solutions (Naylor & Pande, 1981).
In a simplified upper bound solution (in two dhnensions) the continuum is converted to a
mechanis\177n consisting of rigid blocks sliding on their contact areasL For a virtual displacement,
the external work done by the external forces is equated to the internal work done by the inter-
nal forces to obtain the unknown collapse load as an upper bound to the true collapse load. Al-
though the actual behaviour is not the same as the assumed mecl\177anisms, in many cases tl\177e

load obtained for an optimised mechanism is very close to the true load.
The stressstrain model for the soil may be idealizedas an elastic-plastic or a rigid-perfectly
plastic m\177terial. In the latter case, the collapse load is the stone as for an elastic-plastic mate-
rial being independent from the Modulus of Elasticity or stress path. Therefore, in this chapter,
a rigid-perfectly plastic model is assumed.Figure 7.1 shows the stress-strain relationships for a

rigid plastic material. Thereis no elastic deformation as the Modulus of Elasticity is infinite.

On the horizontal plateau the material is in the plastic state and the work done by an increment
of plastic strain is dissipated. Thus the plastic strain increments cannot be recovered during
unloading. The term plastic flow describes the deformation behaviour of the material on this
plateau: In a plastic analysis the plastic strains at failure cmanot be determined. However, using

incremental plasticity, relative rates of strains may be evaluated. In a rigid-perfectly plastic

material the yield point is represented by a yield function F in the stress space. The yield func-

tion represents a fixed surface calledthe yield surface. In the case of work hardening and work
261)))
262 Soil Mechanics:BasicConcepts and Engineering Applications

Rigid work hardening

Stress plastic

Rigid perfectly plastic

/
Yield

Rigid work softening

plastic

Strain
Figurc 7. [. Stress-strain bchaviour of rigid plastic material.

softening materials the yieId function is not fixed and changes with the development of the

plastic strains. The rate of plasticstrah\177 is given by theflow rule according:

G
v = dX O__q__
\177 (7.1)

where G is a plastic potential function (Chen & Saleeb, 1982 and 1986) and dL represents a
positive proportionality parameter (plastic multiplier). When G = F, Equation 7.1
represents
the associated flow rule, other'wise it is a non-associated flow rule. For a rigid-perfectly plastic
material G - F. Furthermore, Equation 7.1 implies that the incremental plastic strains are nor-
mal to the yield surface (Figure 7.2) and that the yield surface must be convex. The yield func-
tion for a two-dimensionaI Mohr-Coulomb failure criterion (Equation 4.11) is expressed by:
, \342\200\242
, 2
F = Y'tC\177x,C\177z,'\177xz)\" =
(\177. _0.\177)2 +(2Zxz)2 _[2c,cos,,+ ((\177,
z ],
+ o.x)sm\302\242) (7.2)

In the undrained conditions Equation 7.2 will form the Tresca yield function given by:

p
\1772,\1772 Tangent to the
Yield surface
\177p

Figure ?.2. Yield surface ancl plastic strain rates.)))

 Application of Limit Attatysis to Stability Problems i\177Soil Mechanics 263

21\177xz

R = 2c cosO+ (O\"z+ O-'x)Sine'

For Tresca material:

/...\177--f--.-.,..,...\177

\177z-\177x
\302\260z-\177x\302\260r

\177
Figure 7.3. Tresca and Mohr-Coulornb yield criteria.

F F(cYx'cYz\"\177xz)
=
(\177z -ox )2 +(21;xr. )2 -(2cu )2 (7.3)
The yield surface is defined 0. In terms of the qua\177atities (o-',, - O\"x) or (\177: - \1770 and
by F =

2Zx:, each of these functions represents a circle of radius 2c\" cost\" + (o\"z- tr'x) sin0' or 2c\177, as
shown in Figure 7.3. Both yield functions can be generalized and expressed in three-
dimensional stress space.In this chapter we will consider only plane strain loading. For appro-
priate three-dimensional yield surfaces reference may be made to Chen & Baladi (1985).
Lower a\177ad upper bound solutions may be obtained using finite elements and linear pro-
gramming (Turgeman & Pastor, 1982; Aysen, 1987; Sloan, 1988 and 1989). In a lower bound
solution, the need for linear programming results from discretisation of the yieId criterion into
linear segments and the mathematical expression of a safe load, which creates inequality con-
straints rather than equalities. Equality constraints are due to the equilibrium and stress bound-

ary conditions. In an upper bou,\177d solution inequality constraints are needed to define that the
rate of work dissipated during plastic deformation or plastic flow is positive, and that the ma-
terial cannot generate energy. Equality constraints arise from satisfying the yield criterion and
boundary displacement conditions, in both cases the collapse load is expressed as a linear
function of the relevant stresses, which is treated as the objective function of the linear pro-

gramming process.

7.2 LOWER BOUND SOLUTION

7.2.l Statically admissible stress field

The lower bound theorem of classical plasticity states that any statically admissible stress field
will provide a lower bound on the limit Ioad. This lower bound limit load is a safe load and the)))
264 Soil Mechanics: Basic Concepts and Engineering Applications

failure of the material will not occur under this load. The necessary conditions required for a
lower bound solution are (Chen, 1975):
1, The stressfield must satisfy equilibrium everywhere within the domain of the problem,
2, The stress field must satisfy the specified stress boundary conditions.
3. The stressfield must nowhere violate the yield criterion.
admissible stress field may not necessarily
A statically represent the actual stresses of the
true solution and include stress discontinuities. There are various ways to construct an admis-
sine stress field. An analytical soIution reIies on reasonable simplifying assumptions and
gives a conservative lower bound load (Mulhaus, 1985). Alternatively, the powerfuI finite
element method can be used to predict the collapse loads m complex stability problems (Nay-
lor & Pande, 1981).A numerical solution due to numerous equality and inequality constraints
needs to be supported by a reliable algorithm to locate the optimal solution on a workstation or

microcomputer.
An alternative method is to
construct the stress field by commencing from a specified
boundary and
moving one zone to from
another zone that is separated by stress discontinuities
in the hope that the stress field will satisfy the numerous boundary conditions of the problem.
This method needs a lot of practice and ingenuity and has been applied to a variety of tunnel
problems (Davis et al., 1980).It is, however, only really useful in demonstrating the applica-
tions of lower bound solutions rather than in the search for an optimal answer.
The third condition of the lower bound theorem, where the yield criterio,\177 is not violated, is
expressed by the foIIowing inequality:
F _< 0 (7.4)

where F is the yield function. This means the state of stress at each point must lie inside the
yield surface, and creates non-linear constraints in terms of the unknown stress components.

7.2.2 Stress discontinuity

Consider the state of stress at point C the comer of the strip footing shown in Figure 7.4. The
two states of stressimmediately to the right and left of this point show a rotation of 90
<'
in the

Figure 7.4. Stress discominuities under the comer of a s\177rip footing.)))

 Application of Limit Analysis to Stability Problems in Soil Mechanics 265

State of stress State of stress

at point 1 at point 2
Figure 7.5. States of slress at a point on a discontinuity.

principal stress directions. The magnitudes of the principal stresses are alsodifferent due to the
different stress boundary conditions. In order to facilitate the construction of an analytical or a
numerical stress field about this point, the adjacent area is divided into a number of fan type
zones. Thus, the point C is represented by 8 points, each with a different state of stress. Due to
the geometry of the problem it is often necessary to insert stress discontinuities within the de-
forming mass in order to satisfy the boundary conditions (Parry, 1995). The states of stress
within the zones are connected to each other through stress discontinuities, each having a zero
(physical) thickness. Equilibrium requires that the \177ormal and shear stresses along each dis-
continuity, calculated from the neighbouring zones, be equal. However, in the theory of plas-
ticity only continuity in the direction
stress discontinuity is required. Figure
normal to the 7.5
shows the geometrical expression of equilibrium at a stress discontinuity. The coordinatesof
the intersection points of the Mohr's circles (Point Tot V) are the normal and shear stressesat
the stress discontinuity. If the state of stressat point 1 is known, then there will be an infinite
number of stressstates at point 2 (or infinite number of circles centred on the or-axis and pass-
ing through r and l0 to satisfy equilibrium of the zones as well as the stress discontinuity.
However, if it assumed that circle 2 satisfiesthe condition F = 0 then there will be only two
solutions. If the condition F = 0 is imposed on both circles then, with the specified state of
stress point 1, there
at will be only one possible state for point 2. In this case the following
equations can be derived from the geometry of Figure 7.6 (Atkinson, 1993; Parry, 1995):

+\302\260x2
\302\260'z2
+c'cot\177'
2 Oc 2 sin(\177+\177.)
c' > 0 (7.5)
+\177rxl
\177J\177i Oq sin(to-\177.)
+c'cot\177)'
2

(Y':2 + (\177x2 _ 0C2 _ sin(t0 + \177)

(7.6)
O'il + O'xl Oq sin(to-\177.)
sinL
sinto =-- (7.7))))
sin\177'
266 Soil Mechanics: Basic Concepts and Engineering Applications

Figure 7.6. Stress discontinuity satisfying Mohr-Coulomb failure criterioa.

where \1770and \177are defined in Figure 7,6 and the stresses are in accordance with the sign con-
'\177ention of Section 4.2.1 or Figure 4.i, Rotation of the principal stress directions due to the
stress discontinuity is defined by \177 that is given by:
- \302\260
- 2\177o
= 2q0:\177 2q0i _ 180.0 = 90
\302\260
_ \177o
\177' (7.8)
2 2

For cu, \177u

= 0 soil all the Mohr's circles ha'\177e identical radii and:

+ q02 = 90 \302\260 (7.9)

q01

Example 7.!
\302\260
A stress discontinuity in a two-dimensional stress field has an angle of 60 with the x-axis,
The state of stressat point ! at one side of the stress discontinuity =
is: Cr'x 100 kPa, o\"; = 200
kPa and -ra: = 50 kPa. Determine: (a) the principal stresses and their direction at point I ; check
if the state of stress satisfies the failure criterion (F = 0), (b) the mag\177fitudes of the shear and
normal stresses along the stress discontinuity, (c) the principal stresses and their directio\177 at
point 2 on the other side of the stress discontinuity that has the same coordinates as point 1 and
satisfies the Mohr-Coulomb failure criterion, (N\177te that if the condition of F < 0 appliesat
point 1, there will be two sets of answers for point 2. Consider the case that has the higher

principal stresses this is not a rule but just a choice for this example). (d) The Cartesian
stresses at point 2. The effective shear strength parameters are c' = 0, \177'= 30 \177'.

Solution:

(a) From Equation 7.2:

F = (200.0- 100.0)
2 \177- \177
+ (2 \303\22750.0) [(200.0 + 100,0) sin 30.0\302\260]
= -2500,0 < 0,
which means point 1 has not failed. Usfitg Equations 4,3 attd 4.4:)))
 Application of Limit Analysis to Stability' Problems in Soil Mectvanics 267

= + 100.0)/2 = 150.0kPa, t or t\" = \177/[(200.0 -

2 + 50.0z = 70.7kPa,
s\" (200.0 100,0)/2]
= 150.0 - 70.7 = 79.3kPa, = 150.0 + 70.7 = 220.7 \"kPa.
cr\177 cy\177

The direction of the major principal plane is obtained from Equation 4.6:

0 = = 22.5\302\260
1/2\303\227tan-t[2\303\22750.0/(200.0-100.0)] .

(b) Calculate the normal and shear stresses at the stress discontinuity (Equations 4.1 and 4.2):
or\" =
(200.0 + 100.0)/2 + (200.0- I00.0) / 2\303\227
cos(2
= 168.3
+ 50.0sin(2\303\22760.0 \302\260)
\303\22760.0\302\260) kPa
z = (200.0- 00.0) t / 2 \303\227
si n(250.0cos(2 \303\227
\302\260) 60.0
\303\227 60.0
\302\260
)
-
= 68.3 kPa.

(c) Relationship between o\"! and o'3 at point 2, which is at the state of failure, is according to
' = 2 \302\260 = ,
Equation 4.13: (r I crg tan (45.0 + 30.0o/ 2) 3or
3.

Substituting the calculated values of the normal and shear stresses along the stress discontinu-

ity into Equations 4.7 and 4,8 (stress transformation equations using principal stresses), and
zmting that o'\177 = 3cr'3, two equations with two unknowns 0\"3 and go are obtained, where go is
the angle of stress discontinuity with the o'\177 plane:

o\" = (o\177 + \177) / 2 + (o]

- cy\177) / 2 x cos 2q0 =
2\177 + cos
cy\177 2go = 168.3 kPa,
=
\"\177 - / 2x sin = sin 2q\177= 68.3 kPa.
(\177{ )
\177r\177 2go \177

equations above yields:3cy\177 -673.2cy\177 + 32990.4 = 0,

2
The solution of the two

Considering the higher value: o\"3 = 152.1 kPa, and O,I = 3O,3-456.3 kPa, 2go = 153.3 \302\260.
Taking the sign convention into account it is evident that the cy\177plane makes an angle of:
/ 2 - 60.0 = 16.65
\302\260 \302\260
153.3\302\260 with tim x-axis.
(d) The \177values for thex and z axes are 16.65 and 16.65 + 90 = 106.65 respectively.
\302\260 \302\260 \302\260 \302\260
From
Equations 4.1 and 4.2:

c\177'\177
= (456.3 + 152.1)/2 + (456.3 - 152.1)/2xcos(2x16.65
= 431.3 \302\260) kPa,

cy\177
= (456.3+ 152.1)/2+ (456.3-152.1)/2 x cos(2xJ,06.65\302\260)
=177.1 kPa,
- - 83.5 kPa.
zzv =(456.3-152.1)12xsin(2x16.65 =83.SkPa,
\302\260) \"\302\242,\177,:

Example 7.2
Two states of stress on two sides of a stress discontinuity are defined s'\177
by:
= 100 kPa, and

s'2 = 300 kPa. Both states satisfy the failure criterion. Determine: (a) the magnitudes of the
principal stresses of both states, (b) the rotation of the principal stresses due to stress disconti-
nuity, (c) the shear and normal stresses along the stress discontinuities T and V (Figure 7.6).
- 20 =
c\" \"kPa, q' 30 \302\260.

Solution:

(a} Referring to Figure 7.6 calculate the radius of eachMohr'scircle:

\302\260 \302\260
sl) x
= xsin = sin = (20.0xcot 30.0 + 100.0)sin 30.0 = 67.3 kPa,
R\177 Oq \177\" (c'cot\177'+ g\177'
\302\260 \302\260
R\177_
= Oc
2 x sin =
\177\" (20.0 x cot 30.0 + 300.0)sin 30.0 = 167.3 \"kPa.

cy'\1771
= 100.0+67.3-167.3 kPa, cy\177
= 100.0-67.3 = 32.7 kPa.
cyi2
= 300.0+167.3 = 467.3 \"l\177a.cr\177,2
= 300.0-167.3 = 132.7\"k_Pa <
cyll
= 167.3 kPa.

Thus the presence of a stress discontinuity is confirmed.)))

268 Soil Mechanics: BasicConcepts and Engineering Applications

(b) Using Equation ?2: sine) = sinX/sin \177'

= sin X/sin 30.0 \302\260
= 2sinX,

:,/1- :,/1- 4sin2 X --,h- - cos 2

:,/4cos2

From Equation 7.5:

300.0 + 20.0xcot 30.0
\302\260
_ = sin(\1770 + %) _ sin tocos\177. + cosoasin %
\302\260
2.4854
100.0+ 20.0xcot30.0 sin(to-X) sintocos\177. -cos\1770sinX

Reptacing sine) and costo in the equation above we obtain an equation in terms of cosX only,
=
from which: X =16.8 =. Thus sino\177 = 2.0 \303\227
sirtL = 2.0 \303\227
sinl6.8 = 0.5?8, =
\1770 35.3 \302\260.

Using the geometry of Figure ?.6; for the stress discontinuity T:

= (03+ X)/2 = (35.3\302\260
+16.8\302\260)/2 = 26.05 \302\260,
q)l
= [(to + ;k) + (180.0= - 26o)]t 2 = [(35.3= + i6.8 =) + (180.0 \302\260
- 2 \303\227
35.3=)] / 2 = 80.75 \302\260,
q)2
= 80.75 \302\260
- \302\260
Tl\177e angle between Oa32 and o\"31 26.05 =54.7 \302\260,

(c) For the stress discontinuity T:

or' = (I67.3 + 32.7)t\"2 + (167.3- 32.7) / 2\303\227cos(2\303\22726.05 =
\302\260)141.3 kPa.
z= (167.3- 32.7)/2\303\227 sin(2 26.05
\303\227 =) =53.1 kPa.
s=
For the stress discontinuity V: o 141.3kPa, z =- 53.1 k_Pa.

7.2.3 Ultimate bearing capacity of a strip footing in Tresca material

The ultimate bearing capacity of a strip footing (Chapter 10) is calculated using Terzaghi's
equation:

q, = c'Nc + qNq +
0.5yBN,/ (?. 10)

in which c' is the effective cohesion,q is the surcharge at the footing level (q = ? D when this
level is lowered to the depth D), B is the width of the footing, Ne, Nq and N-f are Terzaghi's
bearing capacity factors that are functions of the effective internal friction angle

r
C

Figure 7.7. Statically admissible stress field under a strip footing.)))

 Application of Limit Analysis to Stability Problems in Soil Mechanics 269

The bearing may be estimated using

capacity factors the lower bound theorem and by means
of the fan-type stress all of which
discontinuities
pass through the corner of the strip footing. A
classic example of a statically admissible stress field comprising the two zones 1 and 2 is
shown in Figure 7.7. The directions of the principal stresses on both sides of the vertical stress
discontinuity CA comply with the stress boundary conditions. In undrained conditions the fol-
lowing values for the principal stresses satisfy all of the conditions of a lower bound solution:

zone 1 : 0.3 = 5'z, \1771 = 0.3 + 2% = yz + 2c,\177, zone

2:0.3
= \177Z+ 2c u , \1771
=
0\"3
+ 2Cu = yZ + 4c
u .

where y is the unit weight of the soil and z is the depth of the point. In the stress zone 2 and

under the strip footing z = 0 and \1773

=
2cu, qu
= =
\177Yl 4cu \342\200\242

It can be seen that the ultimate bearing capacity of a strip footing in undrained conditions is
independent of the magnitude of the unit weight: N\177,
= 0 and Nc = 4. The value is well
of Nc
below the theoretical value of Nc = (2 + 70
and therefore the number = 5.14 of stress disconti-
nuities has to be increased. It is convenient
to assume equal rotation of the principal stress di-
rectionsacrosseach stress discontinuity. Thus the angles between the stress discontinuities and
the state of stress in each zone are calculated in a manner to ensure equal rotation of the prin-

cipal stress axes.It is also convenient to neglect the weight of the material, as this will reduce
the computations and lead to a direct estimate of the value of N\177..

Example 7.3

Figure 7.8 shows a stripwith 3 stress

footing discontinuities
on c,,, 0, = 0
making 4 stress
zones. It is required bearing capacity (q,) of the footing
to c\177lculate the ultimate and the stress
field within each zone. Assume that the footing is smooth and that the material is weightless.
The stress field within each zone is constant and F = 0, and therefore: \177J\177 - =
c\1773 2cu.

Solution:

stressin 0,crl = 2%.

The state of zone 1 is: =
0.3
The rotation of the principal stress directions acrosseach discontinuity is 90.0* / 3 = 30,0\302\260.

B/2
\" \177
I\177

Figure 7.8, Example 7.3.)))

270 Soil Mechanics: Basic Concepts and Engineering Applications

The angles of the major principal planes in zones 1 and 2 with the stress discontinuity CA are
= 30.0 \302\260
\177 and q\1772respectively, Thus: \177q= \302\242P2-
q01 and from Equation 7.9:
+ q0\177 = 90 \302\260. for and = 30
\302\242P2 Solving \302\242P2 91: \302\242Pl

the stress discontinuity CA are calculated from the state of stress in zone 1:

0.CA
= +
(\302\260'1or3)/2
+ (0.1 - 0.3)/ 2\303\227cos2q) 1 = (2c u + 0)/2 + (2cu -
0)/2\303\227cos(2\303\22730.0
=
\302\260)1.5c u
= = (2c = 0.866c
\302\260)
(0.1 -cr3)/2\303\227sin
\"\177\302\242A 2q\177 u -0)/2\303\227sin(2\303\22730.0 u-

To calculate the principal stresses in zone 2, the normal and shear stresses on the stress discon-
tinuity CA are expressed in terms of the principal stresses in zone 2:
= 1,5c -- -
0.CA u ((Yl + \302\2603)I2 + (\302\260'10.3) / 2xc\302\260s2q\1772
= (0.t -
ZCA
=
0.866c, 0.3)/2X sin 2q\177
2

=2c., = 60.0 \302\260 state

Noting that 0.1 -0.3 and q\1772 , the stress at zone 2 is given by:
= =
0.3 c,,0.1 3c\177,.

Performing a similar calculation and relating the states of stress betweet\177 zones 2 and 3 and be-
tween zones 3 and 4 we find:
zone 3: a.2 = 30.0 \177
0
3
=
2c,\177. 0.1
=
4Cu, zo\177e 4:ct.3 = 30.0 \302\260,
0.3
=
3c\177, 0.t
=
qu
= 5Cu, and
0-4= 60.0 where
\302\260, qu is less than 5.14

Get\177eral equation for Ne in undrained conditions. By considering the geometry of Mohr's cir-
cles of Figure 7.9, the following general equation for q, can be obtained:
= + nc u sin W) = 2c (1 + n sin W)
q\177, 2(c,, u

N,. = 2(1+ nsin \177) (7.11)

where n is the total number of stress discontinuities, and \177 is the rotation of the principal stress
- 10,
directions per one stressdiscontinuity. For example for n gt
= 90 \302\260
/ 10 = 9\302\260:

= 2(1+ 10sin9.0 = 5.13.

\302\260)
N,.

CuNc
--\177 \177_1
I

cu 2cusin\177

Figure 7.9. General solution for N,. in undrained conditions and Example 7\1773\303\267)))
 Applicatiott of Limit Analysis to Stability Problems in Soil Mechanics 271

7.2.4 Ultimate bearing capaci\177- of a strip footing in a Mohr-Coulomb material

For c' = 0 and for the single stress discontinuity of Figure 7.7, with the surcharge q applied on
the ground surface in zone !, stressesin both zones are obtained using Equation 4.13 after
considering \177he continuity of normal stress on the stress discontinuity:
zonel:cr 3=q,\177=qtan-(45 +\177/2),
zone2: (r\177 = q tan 2 (45 \302\260
+ \177 / 2), \177] - qu = q tan
4.
(45
\177
+ \177,/2), therefore:

= tan4(45 \302\260
+ \177'/2)
Nq
The states of stress for two stressdiscontinuities are surrunarized in Figures 7.10 and 7.1i.
From the geometry of Figure 7.i I can be seen that the rotations
it of the principal stresses due
to each discontinuity are identical and equa] to. The total rotation is 90\302\260,
to 90 \302\260- and therefore
is 90 / 2 - 45\302\260.
\302\260
90 - m = 45\302\260.
\302\260
the rotation due to each discontinuity Consequently: to = 45 \302\260.

The angle X is calculated from Equation 7.7:

-\177
sin to = sin 45\302\260
= sin X / sin
\177'--> \177.= sin
\302\260
(sin 45 x sin

Using the geometry of Figure 7. \1770the direction of the major principa] stressescan be found:

= = = =
q0\177 90\302\260-\177zt, \302\24302\177
t35\302\260-ctt, q022 135\302\260-(z\177-c\177 2, \1773 c\177,\177.

As all stress circles are tangent to the failure envelope: the directions above can also be de-
fined from Figure 7.11:
= 22.5\302\260+0.5X, = 67.5\302\260
+0.5\177., = +0.5\177.,
\1770\177 q02\177 \17702\177 22.5\302\260 q)3 =67-5\302\260+05X'

By equating the two sets of angles given above, the unknown values of \177zl, or2 and or3 can be
found as follows:

c\177 -
= 90 \302\260
0.5(6o + X) = 67.5 - 0.5X. c%
\302\260 - to = 45 \302\260
, c\177 = 90 \302\260
- 0.5(6o - X)
= 67.5 \302\260
+ 0.5X.
1 3

Applying Equation 7.6 for successivepairs of stress states of (I,2) and (2,3) generates the
equations necessary to calculate the principal stresses in stress zone 3 which is the requirement
of the problem. Note that the sum of the Cartesian normal stresses to be used in Equation 7.6
is equa\177 to the sum of the principal stresses. The states of stressin the three zones above are:

Figure 7.10. Strip footing with two stress discontinuities.

L)))
272 SoilMechanics:BasicConcepts and Engineering Applications

Figure 7.1 !. States of stress under a smooth strip footing with two stress discontinuities.

zone 1:O\"3 =
' ' ' \303\227tan \177 =
q'0.l =0\"3 2(4 5\302\260+q'/2) q tan2(45\302\260 +\177'/2),

zone 2: --
q sin(o) + )\177)/ =
sin(co --,\177), 0\"\177 q sin(o) + \177k)/ sin(o)
- \177k)tan 2 (45 \302\260
+ \177,/2),
0\"\177
- 2 = q[sin(co + )\177)/ sin(o)- 2 2 \302\260
zone 3: 0.\177
=
q[sin(o) + \177k)/sin(o) \177k)] , 0.] \177.)] tan (45 + \177,/2).
Thus, for two stress discontinuities the parameter Nq is given by:
= [sin(o)+\177.)/ sin(co- 2 \302\260
tan (45 + (/2)
Nq \177,)]2

For n stress discontinuities with n + 1 stress zones\177 the bearing capacity factor Nq
is given by:
= [sin(o)+.k)l sin(co-;k)] n 2 \302\260
tan (45 + @, / 2) (7.12)
Nq
When n approaches infinity the parameter Nq approaches the value obtained from the analyti-
cal method (Chapter 10, Equations 10.2):
= 2 o
Nq exp(r\177 tan \177') tan (45 + \177,/2)

In soil, Nc can be calculated from Figure 7.11 (with n approaching

a c',O' infinity) by moving
the vertical axis z to the right until it becomes tangent to the first circle representing 0%
= 0.
This is equivalent to reducing the o' 1 of the last circle or qu by the amount: q = c' \303\227
cotep'.

Consequently, an identicaI expression to Equations 10.2 is obtained:

- -
cot(p'(Nq 1)
N
c

With the number of stress discontinuities approaching infinity, the magnitudes of at and \177+1
are calculated using the method above with c\302\242\177
replacing :
C\177n\303\267\177

Forn --,,0% co=90 \302\260, =

and \177v 0'. Thus: cq = 45\302\260-0'/2, ctn+\177 =45\302\260+\177'/2.

Example 7.4

A strip footing of width 2 m is located 1.5m below the ground surface. The soil properties are:
c' = 60 \"kPa, \177' y
=
= 23 \302\260. 18.5 kN/m 3. Calculatethe ultimate load per metre of strip footing by
ignoring the weight of \177he soil beneath the footing level.)))
 Application of Limit Analysis to Stability Problems in Soil Mechanics 273

Solution:
q= 18.5kN/m3 \303\227
1.5 m =27.7 kPa.
= \302\260 = \302\260 = 8.7.
Nq
exp(\177 tan \177') tan 2(45.0 + \302\2420\"/2) tan
exp(\177\303\227 23.0 tan2(45.0
\302\260) + 23.0\302\260/2)

Nc = -
1) = cot23.0\302\260(8.7
-
l) = 18.1.
cot\177'(Nq

c+
= c'N = 60.0\303\22718.1 + 27.7\303\2278.7 -- 1327.0 kPa.
q\177, qNq

Q,\177(per metre of strip) = 1327.0\303\227

2.0 1.0 = 2654.0
\303\227 kN.

7.2.5 Stability of a vertical cut h\177Tresca and Mohr-Coulomb materials

Figure 7.12 illustrates the plane strain vertical cut under consideration. The soilis assumed to
be either a Tresca or a Mohr-Coulomb material. For fixed values of H, failure of the cut may
be achieved by increasing either the vertical boundary loading q or the unit weight y. For the
is\177)tropic conditions the stability of the platte strain cut is a function
assumed, of the dimen-
sionless parameters yH/(c\177, and q/(c u or c') called the stability
or c') number and the load pa-
rameter respectively. The stability number is determined assuming the load parameter is zero;
whilst the load parameter is calculated assuming the stability number is zero (e.g. weightless
material). The results may be combit\177ed to illustrate the real situation. A statically admissible
stress field with two stress discontinuities is shown in Figure 7.12. In the stress zone I the
normal stress in the vertical direction is approximated by yz + q, while the horizontal normal
stress is zero. Both stresses represent the major and minor principal stresses, as there are no

shear stresses on the ground surface and the vertical face of the wall. Fan type stress disconti-
nuities in this zone cannot be constructed,as there is no rotation of the principal stresses be-

, y z+q

Z\177

)+q

y
Y (H+zO+q

(H+\177

Figure 7.12. Statically admissible stress field for a vertical cut with two stress discontinuities.)))
274 Soil Mechanics: Basic Concepts and Engineering Applications

tween the horizontal and vertical The stress field in stress zone 2 is hydrostatic
boundaries. in
nature because the vertical and stresses are assumed to be equal.In stress zone 3
lateral normal
\302\260
the vertical normal stress is approximated by yzt and the principal stresses have rotated 90
when compared to those in stress zone 1. The stress field (either effectiveor total) in the three

zones above may be summarized as follows:

zone l:\177J 3 =0,or I --yz +q, zone 2: \1773 =y(H +Zl)+q, cr t =y(H +z\177)+q,
zone 3:or3= Yzl, crl = + zl )
\"\302\242(H + q.
The abovestressfield satisfies equilibrium in the three stress zonesand at the two stress dis-
continuities OAOB. Next, the stress zone 1 is assumed to be on tlxe verge of failure. The
and
major principal stress varies linearly with depth and becomes a maximum at depth H. For the
failure of the toe in a Tresca material we need to enforcethe condition: ch
- or3 = 2c,. Hence
yH + q
- O =
2c, .
= 2 --\177 =0
yH/c u q (7.13)

q/c u
= 2 \177
yH /c u =0 (7.14)
For the stability number yH/c u (with the toad
parameter e\177ua[ to zero), Heyman (1973) pre-
sented an analytical solution and 1+ 242 = 3.83 as the lower
with 2-,]\177= 2.83 and upper
bounds respectively. For a Mohr - Coulomb material with c\" = 0, it is impossible to construct a
vertical slope as, on the vertical face of the slope, \1773 = (t = 0. For a c', 0' material, failure
occurs only at the toe of the slope. The relationship between o\"1 and \1773 at the toe is given by
Equation 4.13 (Chapter 4). Replacingo-'\177 and \1771 by the corresponding values defined above,
two identical equations in terms of the stability number and the load parameter are obtained.
\302\260
yH / c\" = 2 tan(45 + \177'! 2) \177 q = 0 (7. i5)

THIc\177

A 1 2\177/-\177

.\177- q/c u
F [3

Fibre 7.13. Stability chart for a vertical cut using the lower bound theorem,)))
 Application of Limit Analysis to Stability Problems in Soil Mecltanics 275

q / c' = 2 tan(45 =0
\302\260
+ 0'/2) -\177 yH / c' (7.16)
The highest lower bounds found for the stability number and Ioad parameter may be related by
a linear relationship, as shown by the line AB in Figure 7.13 (Bottero et al., 1980). This allows
the estimation of a lower bound for one of the parameters when the magnitude of the other pa-
rameter is known.

Example 7.5
Calculate the bearing capacity at the upper ground surface in a vertical cut 10 in high. The soil
propertiesare:c;\177= 100 kPa, qb, = 0, y = I8 kN/m 3.
Solution:
yH/c, = 18.0x10.0/100.0 = 1.8(point D in Figure 7.13); substituting this value into the equa-
tion of the line AB: yH/c (point E in Figure
u=2,f\177-\177f\177q/lOO.O=l.8--\177q=72.7\"kPa
7.13), and N c = 72.7/100.0 =0.73.

7.2.6 Lateral earth pressure iu Tresca and Mohr-Coutomb materials

A vertical section of a smooth retaining wall is illustrated in Figure 7.14. The failure of the
backfill soil may occur by either of two mechanisms depending on tire direction of the wall
displacement. If the wall moves towards the outside and away from the soil, the resulting fail-
ure is called active.A passive occurs if the wall is pushed towards
failure the backfill until the
limiting displacement is achieved. In a smooth retaining wall the shear stress immediately be-
hi,\177d the wall cannot be mobilized and therefore the vertical and horizontal represent the prin-
cipaI stress directions.The normal stress in the vertical is approximated by yz + q where q is
the vertical boundary loading as shown. In the active state this stress representsthe major
principal stress, while in the passive state it is the minor principal stress. U,\177like the vertical
cut where on the vertical boundary only the toe point is at the verge of failure, all the points

immediately behind the wail are assumed to satisfy the yield criterion (F = 0). the direction
As

of the principal stresses does not change along the wall, there is no need for stressdiscontinu-

Wall displacement
Active \177 \177 Passive q

Figure 7.14. Idealized section of a smooth relaining wall subjecled Io .active or passive failure.

k)))
276 Soil Mechanics: Basic Concepts and Engineering Applications

ity to be included. In the state the lateral

active soil pressure pah is equal to the minor principal
stress o\"3 while O'I = yz In the passive state the lateral soil pressure Pph
+ q. is equal to the ma-
jor principal stress o\"t and o'3 =yz + q. Using Equation 4.13 the active and passive soil pres-
sures are given by the following equations. For undrained conditions:

(r3 = (rl
= - = yz + q - 2c (7.17)
Pah 2Cu u

This linear equation is integrated along the wall to give the total active thrust on the wall:

= 2/ 2+ -
Pall yH (q 2c, )H (7.18)

The distribution of the passive pressure on the wall and the total passive thrust are given by:

Pph
=
cYt
=
o3 + 2cu = yz + q + 2% (7.19)

= yH 2
Ppi,
/ 2 + (q + 2c,,)H (7.20)
For a c', \302\242)'
soil the active and passive pressuresand the total thrusts are given by:

Pah
- (yZ + q) tan
2 o_ (45 - 2c' 9'/2) tan(45
-
\302\260
\177'/2) (7.21)

= (yH 2/2) tan 2(45

\177'
- + H[q tan(45 \302\260-
0\"/2)
- 2c'] tan(45 -
\302\260
Pah q\177'/2) \177'/2) (7.22)

- 2 \302\260
pp\177,
(yz + q) tan (45 + qb'/2) + 2c' tan(45\177+ 0'/2) (7.23)

Pph =(YH2/2)tan2(45\302\260+e'/2)+H[qtan(45o+\177'/2)+2c']tan(45\302\260+O \"/2) (7.24)

The lateral soil pressure is linear indicating an increase in active or passive soil pressurewith
depth. In the active state there is a possibility that tensile stress may develop behind the retain-
ing wall to a depth of zo, which is estimated by setting the active pressureequation to zero. As
the soil cannot sustain tensile stress, tension cracks developto the depth zo. In a rough wall
shear stress will develop along the back face of the wall causing rotation of the principal
stresses. The shear stress is either of a cohesiveor frictional nature and causes a reduction in

yz + q
\177\\\\\177

O\"
h

\177xz
atum

Figure 7.15. Active failure of a rough wall.)))

 Application of Limit Analysis to Stability Problems in Soil Mechanics 277

O
- 2c\177

\177\177t
V

Figure 7. I6. Stress circlesof active failure for c,, = 0

\177, soil behind a rough vertical retaining wall.

the lateral soil pressure.To allow for the rotation of the principal stresses, a stress discontinu-
ity is introduced which makes an unknown angle ct with the back face of the wall. In an active
state and in the undrained conditions, the major principal stress in zone 1 (Figure 7.15) is ver-
tical and its magnitude is yz + q. The minor principal stress + q - 2c,. In
in this zone is cr\177= \"\302\242z
the stress zone 2, the normal stress in the vertical oh, altd in the horizontal Pah and the angle
are all unknown. However, the shear stress rx: is known and is equal to the cohesion Cw devel-
oped on the wall and the soil contact surface. Assuming that the stress field iu zone 2 meets
the requirements for failure, the geometry of the stress circlesshown in Figure 7.16 is used to
find all unknown parameters. The rotation of the principal stresses due to the stress discontinu-
= - 7.9, \177, q0\177and (P2 can be expressed in terms of the angle
ity is \177 q01 q02.Using Equation
+
\342\200\242
1
= 7\177/2,g/2-
\177\302\2602
ct + \17702 - \177t/2 --\177 (92 = or. Thus:
= - = \177z/2 -\177 -\177 -- n/2-2c\177
\177 q\177 q\177 (7.25)
The angles of the major principal planes of zones I and 2 with the back face of the wall are /
\177t

2 and q0A (Figure 7.16) respectively. Hence:

= 72/2 - - = 0/2
\177 -q0\302\242,
\177t/2 (\177 -\1773)/2 (7.26)
where 0 is definedby:sin 0 = c\177,,/c\177,. Equating Equations 7.25 by 7.26:

=
c\177. rt/4- 0/4 (7.27)
From the geometry of Figure 7.16 the active lateral pressure pah is:
=
Pah (yz + q)-c u -ClC 2 -c u Cos0, or:

Pah
= + q)
(\302\245Z -c (1 + 2sin 13/2 + cos0) (7.28)
u

Integrating the equation above aIong the wall, the total active thrust is expressed by:

Pat,
=
VH
2
/ 2 + [q - C
u (1 + 2 sin 13/2 + cos 13)]H (7,29)
If the number of stress discontinuities is increased from 1 to n, then the term 2sin(13/2)
Equation 7.29 may be replaced by 2nsin(0/2n). However. the improvement (decrease) in the
active pressure is not significant. The passive pressure ca\177 be calculated in a similar manner:)))
278 Soil Mechanics: Basic Concepts and Engineering Applications

Figure 7.17. Stress circles of active failure for c\" = 0,

\177'soil behind a rough vertical retaining wall.

= + q)
(\302\245z + cu (1 + 2 sin 0 / 2 + cos0) (7.30)
pp\177

The total passive thrust is:

Pp\177,
=
yH 2/2+ [q +c\177, (l +2sin0/2 + cos 0)]/-/ (7.31)
The stress circles for a c' = 0, 9' material are shown in Figure 7.17. The shear stress on an
element of soil immediately behind the retaining wall is: \"r\177z=pat, tan6\", where 6' is the friction

angle mobilized on the interface. The active pressure is expressed by:

= Oc
p,\177t\177 2
- R 2 cos 0 = Oc2 - Oc
2 si ridS'cos0 = Oc2 (1 sin 0'cos 0)
-

where the angle 0 is defined in Figure 7.17. To relate 0c2 to Ocl we use Equation 7.6:

sin(to-\177.)(1- sin
= Oc \303\227 \177'cos0), but
P,\177 I
sin(\177o + X)

Oc -
yz + q
- R --
yz + q
- sin \177' --\177 Oc
=
(Yz + q)/(1 + sin \177'), and hence
I 1 Oc\177 I

sin(e\177- L)(l- sin \177'cos0)

=(yz+
P\177I\177 q) =(\302\245z+q)k'\177l\177 (7.32)
siu(e\177 + \177,)(l + sin q\177')

The term k\177 (soil pressure coefficieut iu the horizontal for an active failure) is given by:

- sin(e,\177- X)(I sin q\177'\302\242os0)

k\177t\177 (7.33)
sin(to + \177,)(1 + siu 0')

Using a similar approach, the soil pressurecoefficient for the passive case can be obtained:

- siu(to + \177.)(1 + sin 0'cos O)

(7.34))))
k\177/,
sin(c0- \177)(1 sin 0')
 Application of Limit Analys(s to Stab(lity Problems (n Soil Mechanics 279

The total active or passive thrusts are the integral of the soil pressure along the wall:
2
P,\177I,
=
(yH / 2 + qH )ka\177, (7.35)

= (yH 2 / 2 + (7.36)
Ppt, qH)kp\177,

The angle 0 can be expressedin terms of 6' and (\177'by using Equation 7.7. Introducing the an-
gle to' defined in Figure 7.17, then:

sin w' = sin(0 + 8') = sil\1778'/sin \177' (7.37)

For the passive case this equation has the following form:
sin 0)' = sin(0- 8') = sinf'/sin \177)\" (7,38)

Note that the angle 0 / 2 represents tbe rotation of tl}e principal stresses \"q (Equation 7.26) due
to the stress discontinuity, At the same time, the rotation of the principal stresses is equal to
\302\260
\177
= 90 -to (Equation 7,8). Thus, by equating the two values above, the angle to can be calcu-
lated, It can be shown that increasing of number of stress discontinuities will not significantly
improve the solution,

Example 7.6

Calculate the total horizontal thrust on a retai\177fing wall of height 10 \177nin both the active and

passive cases. The soil properties are: c' = 0, \177'= 30 6' =

\302\260, and )' = 18
20 \302\260, kN/m 3.

Solution:

For the active state use Equation 7.37:

sin((3 + 6') = sin(0 + 20.0\302\260)

= sin fi'/sin = sin 20.0\302\260 \302\260
= = 23.16
\177)' / sin 30.0 0,684 --\177 0 \302\260,

0 / 2 = rotation of the principal = 23.16 \177 = 11.58 \302\260,

stresses / 2
= \302\260
\177
= 0/2 23.16\302\260/2 = 90.0\302\260- to---\177 0\177= 78.42 sin
\302\260, 78.42 = sin X/sin 30.0\177--)
X = 29.33 \302\260,
\302\260
- \302\260
o; = ((t) X)/2
= (78.42 29.33\302\260)/2
= 24.54 . From Equation 7.33:
\302\260
- - sin \302\260
sin(78.42 29.33\302\260)(I .30.0 cos
\303\227 23.16 \302\260)
= 0,2858\177
\302\260
sin(78.42 + 29.33\302\260)(i + sin 30,0 \302\260)

This value is slightly greater than 0.2794, which is based on the analytical approact\177 using tlae
limit equilibrium method (Chapter 8).
2 = 257.2 kN.
Pa\177
=
(yH / 2 + qH )kah
=
(18.0 10.02
\303\227 / 2)0.2858

For the passive state use Equation 7.38:

sin(0
- 6\") = sin(0 - 20.0\302\260)
= sin 6\"/sin 0' = sin 20.0\302\260/sin 30.0 = 0.684 \302\260
--\177 0 = 63.16 \302\260,
\302\260
0 / 2 = rotation of the principal stresses = =
63.16 / 2 31,58 \302\260,
= 0/2 = 63.16\302\260/2 = 90.0 \302\260
-co ---) \177= 58,42 sin
\302\260, 58.42
\302\260
= sil\177X/sin 30.0
\302\260
\177 X = 25.21 \302\260,
\177
\302\260
(z = (\177o+ X)/2 = (58.42 +25.2l\302\260)/2 =41.81 \302\260,

\302\260
sin(58.42 + 25.21\302\260)(1 + sin 30.0\302\260 cos 63.16
\303\227 \302\260)
= = 4.448.)))
kpl \177
sin(58.42\302\260- 25.21\302\260)(1-sin30.0\302\260)
280 Soil Mechanics: BasicConcepts and Engineering Applications

This is much smaller than the value of 5.7372obtained form the limit equilibrium method
(Chapter 8), and theretbre the number of discontinuities must be increased.

=
(yH
2
/2+ = (18.0\303\22710.02/2)4.4483 = 4003.4 kN.
Pph qH)k#,

7.3 UPPERBOUNDSOLUTION

7.3.1 Kinematicalty admissible vetocity field

The upper bound theorem of limit analysis defines a velocity field(or plastic strain rate field),
which is said to be kinematicaIly admissible if the following requirements are met:
1. The velocity field must satisfy strain compatibility.
2. The yield function expressed by Equations 7.2 or 7.3 must be satisfied.
3. The stress-strain relationship in the tbnn of Equation 7.1 must be satisfied within the field.
4. The boundary veIocities must be satisfied.
There is no requirement for the equilibrium condition, and the velocity field may contain ve-
locity discontinuities. A velocity discontinuity is needed to distinguish between parts of the
material that may remain rigid while the other parts are on the verge of failure or have already
failed. This could occur at both the boundaries of the domain of the problem as well as in its
interior. By applying the virtual work theorem, which equates the external work done by

boundary and gravity loading to the internal work dissipated by plastic straining within the
material and by sliding along the velocity discontinuities, an upper bound limit for the collapse
of the rigid body can be found. This means that a stress field associated with the kinematic ve-
locity field is created to evaIuate the stresses in the interior of the domain and aIso at its
boundaries. From this the unknown colIapse load that may rise from the boundary stresses

(e.g. the ultimate bearing capacity of a strip footing) or from the gravity related parameter (e.g.
yH / c' in the stability of s\177opes) can be found. Since the stress field does not necessarily sat-
isfy the equilibrium conditions, the upper bound load is an unsafe load and mathematically de-
fines an external approach to the yield surface. The ten\177 upper bound does not mean that the
collapse load is greater than the lower bound collapse load as this depends on the mechanism
of the collapse. An upper bound solutiou is an external approach to the yield surface while a
lower bound solution implies an internal approach. For example, in the case of a strip footing,
the upper bound value for the ultimate bearing capacity is higher than the lower bound value.
In the active faiIure of a retaining structure, the lower bound value is greater than the upper
bound value. This simply implies that if there is a smaller resistance from the retaining wall
against the applied pressure from the soil, then the structure will fail. As was mentioned ear-
lier, the exact solution remains in between the upper at\177d lower bound solutions.

7.3.2 Velocity discontinuity and the concept of two-dimensional collapse mechanisms

In a rigid plastic body some parts may be deforming continuously while other parts may not be
deforming at all, and this gives rise to the strain rate or velocity discontinuity (Naylor & Pande,)))
 Application of Limft Analysis to Stability Problems in Soil Mechanics 281

t98 l), The velocity normal to discontinuity must be continuous a gap will develop
otherwise
and parts of
body will penetrate into each other.velocity Adiscontinuity is also referred to as
a slip surface (Atkinson, 1993) if a rigid body is modelled by a mechanism of rigid blocks
sliding on their contact surfaces with constant velocity. Although the external work is dissi-
pated by general plastic yielding and sliding along the discontinuities, in a collapse mechanism
it is solely dissipated along the velocity discontinuities. This approach is commonly appliedin
plane strain conditions and the dissipated work is minimized by optimising the geometry of
the mechanism. Figure 7.18(a) is an example of a rigid block mechanism used to investigate
the stability of a plane strain
slope. The mechanism satisfies the velocity compatibility condi-
tion and is geometrically defined by relevant angles or lengths. Each velocity vector makes an

angle of \177' with the direction of the discontinuity. This does not satisfy the continuity of a
velocity discontinuity, as defined in the theory of plasticity, as in a collapse mechanism a jump
in normal and tangential velocity may occur. For a Tresca material with 0u = 0 there is no

jump in the normal velocity and the velocity vector is parallel to the direction of the disconti-
nuity. For a Mohr-Coulomb material, the \177' angle assumed between the velocity vector and the
discontinuity facilitates the f/ow rule conditions. In order to establish a relationship between
the magnitudes of the velocities and thereby satisfy the compatibility of the displacements, a
diagram of velocities is drawn (Figure 7.18(b)). This diagram is called a displacementor ve-
locity diagram. To construct this diagram for the mechanism shown in Figure 7.18(a), three
vectors parallel to the relevant velocities are drawn. The end points of these vectorsare con-
nected to each other to represent the relative velocities on the discontinuities. Assuming that
one of these velocities is equal to a specified value (e.g, Vl = 1 unit) we can calculate the other
velocities from the geometry of the displacement diagram. The work done by the external
loading includes tl\177e boundary loading q and the weight of each block. The algebraticvalue of
the work is equal to the magnitude of the force multiplied by the component of the velocity in
the direction of the force. If this component and the load are at the same direction then the
work may be assumed positive,otherwise it is negative. The sign of the dissipated work is al-
ways positive and is due solely to cohesion. The resultant of the normal stress o-' and the shear

(a) to bc

(b)

Figure 7.18, (a) Mechanism of rigid blocks, (b) displ,\177cement diagram.)))

282 Soil Mechanics: Basic Conceptsand Engbwerfng Applications

FI

Figure 7.19. The concept of dissipated work along a velocity discontinui|y or a slip surface.

stress component \177 tanqb' (R in Figure 7.19) is perpendicular to the velocity discontinuity
- as
a result the internal work due to this stress is zero. Referring to Figure 7.19 and noting that the
incremental displacement is inclined to the slip surface at an angle \177', the internal work dissi-
pated along the velocity discontinuity of \177ength I may be expressed as:
Ei =
J0c vcos0
dt (7.39)
Equation 7.39 must be applied for each discontinuity and algebraically combined to represent
the overall internal work dissipated in the collapse mechanism.

7.3.3 Application of two-dimensional collapse mechanisms to ascertain bearing capacity

Fan type velocity discontinuities radiate from the corner of a strip footing and satisfy compati-
bility of the displacements. The following example illustrates the evaluation of an upper bound

value for the bearing capacity factor Nc of a Tresca material.

Example 7.7
For the collapse mechanism given in Figure 7.20(a), calculate the upper bound value for the
ultimate bearing capacity of the strip footing if the soil is undrained.
Solution:

From the geometry of the problem: ab = be = ec = cd=\177/-\177-B

/ 2 = 0.707B, bc = B,
= w = w = abxeb/2xy = =
w\177 2 3 (0.707B\303\227O.707B/2)_x\302\245
0.25B2\302\245.
From the displacement diagram: =
v\177 v3
= 1.0, v =
2 42 = 1.414, v12 = v23 = 1.0.
Calculate the external work due to the weight of each block and q,:
Ee\177
= w x
I cos45.0
v\177\303\227
= 0.177B \177,
\177-
wvx v2x cos90.0 = 0.0,
Eq2
Ee3= w3 x v3 x - cos45.0\177= -
0.177B'y, Ee\177
=
q, B x
\303\227 Vl cos45.0
\303\227
\177
=
0.707Bq,.
Total externaI work =
0.707Bqu. Internal work along each velocity discontinuity:
=
E\177.I ab x cu \303\227
vl \303\227
cos0u
= 0.707Bcu, Ei2 = bcx c, x v2 x cosqb u = 1.414Bcu,
El3
= cdx cu = 0.707Bcu, El4 =
v3 \303\227
\303\227 cos\177u be x c, x v12 x cos0u = 0.707Bcu,
El5
= ce \303\227
G, \303\227
v23 \303\227
cosmic
= 0.707Bc,, total internal work = 4.243Bcu. Equating the external
a\177d internal work: 0.707Bq, = 4.242Bc,, q, = 6.0c,.
Thus, tile upper bound for Nc
= 6.0 < 5.14.)))
 Application of Limft Analysis to Stability Problems in Soil Mechanics 283

Figure 7.20. Example 7.7.

7.3.4 Application of two-dbnensional collapse mechanisms to the stability of slopes

For the mechanism shownin Figure 7.18(a), a simple computer program may be written to
calculate the stability numberor the load parameter. The number of fan angles and their mag-
nitudes (13t, 13:\177,133 .... ) along with parameters \177tt, c\1772and 0t3 can be given as data, whiIe the fi-

nal values ,nay be found through an optimisation procedure. In generaI, a fan of 3 to 4 triangu-
lar bIocks is adequate to yield reasonable values for yH 1 c' or q / c'. For a vertical slope (or
vertical cut), one variabIe mechanism consisting of one triangle yields a reasonable solution.

Table7.l(a). Loadparameter q / c' for a vertical cut with TH / c' = 0.

LB UB UB Slice
\177'(deg.) FEM FDM
FEM FEM Mech. method

0.0 1,93 2.00 2.00 2.00 2,00 2.00

10.0 2.29 2.40 2.40 2.38 2.38 2.38
20.0 2.71 2.93 2.90 2.85 2.85 2.85
30.0 3.24 3.67 3.50 3.46 3.46 3,46
40.0 3.90 4.97 4,50 4.34 4.29 4.30
50.0 4.77 7.70 5.20 5.00 5.49 5.50

Table 7. I(b). Stability number yH / c' for a vertical cut with q / c' = 0.

LB UB UB Slice
FEM FDM
0\"(deg.)
FEM FEM Mech. method

0\1770 3.60 4.00 3.60 3.48 4,00 3.83

10.0 4\17735 5.10 4.37 4.54 4.67 4.66
20.0 5.I6 6.50 5.90 6.46 5,71 5.64
30.0 6.12 8.55 7.37 7.65 6.92 6\17787
40.0 7,24 [3.16 9.56 [0.00 8.57 8,53
50.0 8.62 26.14 18.00 10.98 [0.89)))
284 Soil Mechanics: Basic Concepts and Engineering Applications

.7.70

mechanism
o=

2 \177nd
1 _Slicerneth0d

0
0 2 3\"684 6

q/%
Figure 7.2 I. Stability
\302\260
bounds for a 45 slope in undrained conditions.

Aysen & Loadwick (1995) compared the mechanism solution with the traditional slice
method based on the limit equilibrium (Chapter 9) and available upper and lower bound solu-
tions based on the finite element & Sloan, 1992),Furthermore,
method (Aysen a displacement
finite efement method &
(Carter Balaam, 1990) and an explicit finite difference method (Cun-
dall, 1980 and 1987) were also used to evaluate the mechanism results. Tables 7.1(a) and
7.1 (b) present a detailed comparison for a vertical slope, In general, the difference between the

lower bound and tnechanism solutions is very small and at low values of 0' they are ahnost
identical. At high vaIues of \177b',the upper bound finite eIement solutioll does not yield reliable
answers. Results obtained from the displacement type finite element method and finite differ-
ence method using the fast Lagrangian analysis of continuum (FLAC) are reliable at\177d fall be-
0u =
\302\260
tween the two bounds. Figure 7.21 shows the stability chart for a45 slope with 0 where
the two extreme points of each solution have been connected by a straight line.

Example 7.8
For a 45 \302\260
plane strain slope, compute the upper bound value of q for the collapse mechanism
shown in Figure 7,22(a). The soil properties are: cu
= 100 kPa, = 0, and
t\177u
= t8 kNlm 3.
\302\245

Solution:
\302\260
From the geometry of the problem: hi - 20.0 x sin60.0 = 17.32 m,
h2 = 30.0 - 17.32=12,68 m, l=30.0 + 20.0 / 2=40,0 m, cq =tan-l(12.68/40.0)=17.6
\302\260,
= 45.0 \302\260
\1775 - 17.6 \302\260 ab = bd
= 27.4 \302\260, - 20.0 m, bc = l / cos cz = 40.0 / cos 17.6\302\260
= 41.96 m, t
The weight of eacll block is calculated to be: Wl = 3117.6 kN, w2 = 7376.54 kN.
From the displacement diagram of Figure 7.22(b), vl = 1, v2 = 0.887 and v\1772 = 0,690.
Components of the externaI work:
Eet = w\177\303\227
\302\260
= 31 i7.60 \303\227 \302\260
= 2699.92,
v\177cos30.0 1.0 x cos30.O

Ee2 = v2COS72.4
W2\303\227
\302\260
= 7376,54 x 0-887 X COS72.4
\302\260
= 1978,40,
\302\260
Ee3 = 20,0 COS30.0
q \303\227Vl\303\227 =
17.32q. Hence, the total exten\177al work = 4678,32 + 17.32q.)))
 Application of Limit Analysis to Stability Problems in Soil Mechanics 285

(a)
Figure 7.22. Example 7.8.

Ell
= ab \303\227
cu \303\227 = 20.0 i00.0
\303\227 [.0
\303\227 = 2000,
vlcosOu
Ei2= bc \303\227
c, \303\227
v2cos\177u
= 41.96 100.0
\303\227 \303\227
0.887 = 3721.85,

El2= c, \303\227
bd\303\227 vl2cos\177,
= 20.0 \303\227
100.0 0.690
\303\227 = 1380.00.
Total internal work = 7101.85. Equating the external work and the internal work:
4678.32 + [7.32q = 7101.85,q = 140kPa, q/c, = 140.0 / 100.0 = 1.4.
Note that the given angles in the geometry of the mechanism are not optimised values.

7.3.5 Application of two-dimensionaf collapse mechanisms to retahzing walls

Figure 7.23(a) illustrates a simple upper for the active (solid lines) or pas-
bound mechanism
sive failure of a retaining wall. There may interface of the retaining
be frictio\177 at the
wall and
the backfill, as well as cohesion.It is convenient to assume that a hard stratum occurs below
the base of the retaining wall that is not subject to failure. Whilst the back face of the retaining
wall may or may not be vertical, the upper bound analysis applied to the conditions shown in

Figure 7,23(a) is equally applicable for non-vertical retaining wall. For the active failure case,
the external work due to the weight of the block, the load q, and the horizontal and vertical

components of Pa are:
Eel
=
H(H tano\177/2)y\303\227v a cos(\177)'+ o:), Ee2 = H tano\177\303\227q\303\227
v a cos(\177'+ c\177),

= =
Ee3 -Pa c\302\260s6'va sin(0\177' + ct), Ee4 -Pa sin \177i'va cos(0\177' + cQ.
The internal work is due to the cohesion c\" along the sliding surface of length H/cosc\177 :

E i = (H /cos ot)c'xv a cos 0\". Equating the total external work to the total internal work:

2 tanctcos(\177b'+c\177) tan c\177

cos(0/+ c0 cos \177b'
= c'H
Pa yH \303\267qH (7.40))))
12 sin(8'+ 0' + c\177) sin(6'+ 0'+ c0 cosctsin(8'+ \177'+ a)
286 Soil Mechanics: Basic Conceptsand Engineering Applications

I
I
i

(a)

Figure 7.23. AN>lication of a single variable mechanism to a retaining wall.

In undrained conditions with 6' = 0, Equation 7.40 simplifies to:

Pa
=
7H 2/2 + qH -cull/sin o\177cos ct (7.41)

of Pa (in - \302\260
Setting the derivative terms of o 0 to zero, \177 45 and:

Pa =Y H2 I2+(q- 2cu)H
which is identical to Equation 7.18 derived from the lower bound theorem. Similarly for the
passive case, an equation identical to Equation 7.20 can be obtained, For the case when the ef-
fective cohesion is zero, Ei = 0 and the total thrust is given by:
2
Pa
=
(7H / 2)[tan o:cos(\302\242'+ \177) / sin(6' + + c\177)]
\302\242' (7.42)

A soil pressure coefficient and a lateral component can be defined as follows:

ka = tan + \177)/sin(8' + e' + c\177), = k cosS' (7.43)

o\177cos(O' kah a
For passivefailure of the backfill the corresponding values of kp and kph are:
= tanc\177cos(o\177-O')/stn(c\177-6'-\177'), cos\177' (7.44)
kp kph =kp

Example 7.9

Re-work Example 7.6 using the single variable mechanism of Figure 7.23(a).

Solution:

The optimum value of angle o\177is calculated by trial and error using Equations 7,43 and 7.44,
\302\260
which is 34 and 72
\302\260
for the active and passive statesrespectively (refer to the table belew).

30 33 34 35 36 40
0+2754 0.2791 0.2794 0,2791 0.2784 0.2697
51
\177\302\260 60 65 70 7t 72 73 74 75
kph 62.0744 8.1172 6.3779 5.7826 5.7473 5.7373 5.7530 5.7957 5.8677)))
 Application of Limit Analysis to Stability Problems in Soil Mechanics 287

\302\260
Pah
= (18.0 10.02
\303\227 /2)\303\2270.2794
= 251.5 kN < 257.2kN (LB), Pay
=
Pat\177
tan
\303\227 20.0 = 91.5 kN,

Pa
= \177251'52 +91\"52 = 267.6 k_N.
= ( I80\303\2271002/2)\303\2275.7373 = 5163.6 > 4003.4 \"kN (UB),
Ppl,
= 1879.4
\303\227tan20-0\302\260 \"kN, = \1775163.62 + 1879.42 = 5495.0kN.
Ppv Pph Pp

7.3.6 Application of two-dimensional collapse mechanisms to shallow tunnels

Collapse mecba\177fisms bare been applied to the stability of shallow tunnels in conjunction with
centrifugal tests on soft clay performed at Cambridge University (Mair, 1979). Davis et al.
(1980)obtained reasonable upper bound solutions for circular tunnels (Figures 7.24(a) to
7.24(d)) and tunnel headings (Figure 7.26(a)) using 1 to 3 sliding blocks. The mechanisms in-
cluded local collapse ('not shown) as well as the active total collapse of the section. Stability
ratios were formulated in terms of the geometry of the sliding blocks. Sloan & Aysen (1992)
used a seven variable mechanism ('Figure 7.24(e)) with cohesion increasing linearly with

depth. An application of this mechanism for uniform strength is shown in Table 7.2. Britto &
Kusakabe(1985)applied the upper bound mecha\177fisms to axisymmetric problems with an em-
phasis on simplicity and the ease with which calculations could be carried out. The ability of
the collapse mechanisms in tunnels to capture the results from the numerical based methods is

(a) (b)

(d) (e)

Figure 7.24. Collapsemechanisms for a plane strain shallow circular tunnel in undrained conditions.)))
288 Soil Mechanics: BasicConcepts and Engineering Applications

(b)
Figure 7.25. Collapse mechanisms for a shallow square tunnel in undrained conditions
(Aysen & Sloan, 199Ia).

Table 7.2. (qs- qt) for circular

/ cL\177 tunnel using the collapse mechanism of Figure 7.24(e).

H/D yD / cu = O yD / cu = I yD / cu =2 yD / c. =3

1.0 2.549 1\177366 0.151 - 1,I21

2.0 3,676 1.410 - 0.907 - 3.281
3.0 4.505 1.179 - 2.197 - 5.627
4.0 5.174 0.807 - 3.611 -8.076
5.0 5,745 0.345 - 5.103 - 10.595
6.0 6.246 -0.179 -6,651 - 13.163
7.0 6.697 - 0.750 - 8.24I - 15,770
8.0 7.108 - 1.358 -9.865 - 18.40\177,
9.0 7.486 - 1.995 - 11.516 - 21.062
tO.O 7.838 - 2.657 - 13.t89 - 23\177740

reported by Aysen & Sloan (1991a).The stability ratio may be represented by the dimen-
sionless parameter qs- qr ! c\" or cu where qs and qr are the vertical boundary load and the tun-
nel pressure (reactionof the tunnel lining or an actual hydrostatic pressure) respectively. Fig-
ures 7.25 and 7.26 (Aysen & Sloan (i991c)) illustrate the application of collapse mechanisms
to square tunnels and tunnel headings.

Example 7.10
Using the collapse of Figure 7.25(b) and the following
mechanism data, compute the upper
bound value for (qs - in a square tunnel
qr) ! cu \177nundrained conditions (cu, \177u= 0.0). ct = 60\302\260,
13
=
= 75 \302\260,
c\177 H / B
75 \302\260, = = is the unit weight of the soil,
5, yB / cu 3, where \302\245
Solution:

The geometry is symmetric, and therefore only one half of the mechanism is considered.
Relevant dimensions are: dclsin45.0\302\260 similarly: bc = 0.732B,
= B/sin \302\260 = 0.732B,
75.0 -\177 dc

oc = 0.896B,ob = B, and ab = 5B.The

weights of the blocks: w
I
\177-
5B(0.5B + B)y = 7.5B2y,
w = w = od xsin60.O\302\260xdcl2x\302\245 =
Bxsin60.O\302\260xO.732B/2\303\227y
=
2 3 0.317BZ\302\245.)))
 Application of Limit Analysis to Stability Problems in Soil Mechanics 289

'
Parallel

Parallel\177

Figure 7.26. Collapse mechanism for a plane strain shallow tunnel heading in undrained conditions:
(a) Davis et al. (1980), (b) Aysen & Sloan (199lc).

From the dispIacement diagram shown in Figure 7.27 and assuming Vl

= I:

v2= v3
= 1.155, vt2 = 0.577, v23 0.598.
=

Eel = wl x vl = 7.5B2y x h0 = 7.SB_2y,

\177 2 2
Ee2 \177-w2 x v'\177\303\227
cos30.O = 0.317B_y x h \17755 x cos30.0 \177= 0.3 \1777B
27'
Ee3 = w3 x v\177X cos60.O 0.317BZy x 1. I55 x cos60.0= 0.183 B y,
\177 \177
=

Calculate the external work due to the applied s\177esses:

Ee4 = qs (0.5B + B) x vl = 1.5Bqs,
Ee5 =
- qr x B x v3 sin 60.0 \177= - qtx B x 1.155x sin60.0 = -
\177
Bqt,
Ee6 =-qtxB/2xvl qr,
=\177B/2
Total external work = 8B \177+ 1.5B(qs - q\302\242). Calculate internal work along each discontinuity:
Ell = ab x cu x vlcos\177u = 5B x cux [.0 = 5Bc,,
El2
= bcx cu x v2 cos\177u = 0.732B x c, x 1.155= 0.845Bcu,
El3 = cd x cu x v3cos\177u
= 0.732B x c, x I.155= 0.845Bc,,
&4
= obx cu Vl2COS\177, = B x cu x 0.577 = 0.577Bcu,
x
El5
= oc x cu x v23cos\177u = 0.896B x cu x 0.598 = 0.536Bcu,
Total internal work = 7.803B c,. Equating the external work to the internal work:
8B2y+ 1.5B(qs qr)
- = 7.803B cu \177 (qs- qr) / c, = - 10.8,

V12

Fibre 7.27. Ex\177ple 7.10.)))

290 Soil Mect\177anics: Basic Concepts and Engineering Applications

From the FEM solution using Table 7.4, (qs - qr) / cu = - 11.2.

Example 7.11

A vertical section of a plane strain circular tunnel with a 4 variable upper bound mechanism is
shown in Figure 7.28(a). The magnitudes of the variables are selected as follows: \177= t5 \302\260,

I\177
= \177= 0 = 75 \302\260.
60 \302\260, For the undrained condition find the ratio of q.s/ c, ifyD / cu = 1.0. What
will be the tunnel pressure q\177if the applied boundary stress qs = 4 c\177'?

Solution:

As the section is sy\177etrical about OE, only half of the section is considered.
sl = \177ea ofEDCAF = \177ea EDGO - \177ea OACG - area OAF,
Sl
= OExOG - [OK \177AK 12 + (AK + CG)I2 x(OG- OK)]- 15.0
\177
\177OA
\177
/ 360.0\177; solving the
geomet\177: OE = 7.0 m, AB = 1.0 m, CJ = 1.866m, OC = 2.732 m, CG = \177.932 m,
2
OG = 1.932 m, ED = OG = 1.932 m, AK= 0.966 m, OK = \177.259 m, thus: s\177= 10.844 m .
S(ABC) = S2 = AB\177OBcos30.O\177
\177
/2 + ABxCJ /2 \17760.0
x\177OB2 /360.0\" = 0.842m 2.
The ienglh of the veiocily discontinuities are: CA = CB = 1.932 m, CD = 5.068m.
The velocity of each block is calculatedfrom the displacement dia\177am of Figure 7.28(c):
vi
= 1.0,
v2 = 1.732, v\1772
= 1.0.

Calculation of components of the exte\177al viauaI work:

Eel = w\177\177Vl = s\177\177y \177Vl = 10.844 \177y \1771.0 = 10.844y,
\177
Ee\177
=
w2 x cos30.0 \177= s2 x
\177v2
y
z v2 \177cos30.0
\177
= 0.842 \177y x \177.732 x cos30.0 =
1.263y,
Ee3 = qs x ED \177vl = 1.932 qs, total external work = 12.107y + 1.932 qs.
Calculatioa of componentsof the iute\177al visual work:
Eil = DC x cu = 5.068 x cu x 1.0 = 5.068 cu,
x vtcos0u

El2
= CA x cu x
Vl2COg\177u
= 1.932 x cu x [.0 = 1.932 cu,

Ei3 = CB cux x v2cosO,

= \177.932 x cu x 1.732 = 3.346c,, total internal work = 10.346 c,.
Equating the external work to the internal work: 12.107y + 1.932 qs = 10.346
that: yDlcu = \177.0, yx2.0/c =L0\177y=0.5 c,, thus:
Noting u

,.

\177 \177 15 \177

E\177

D=2.0m \177

(a)

Fibre ?.28. Example 7.] ].)))

 Application of Limit Analysis to Stability Problems in Soil Mechanics 291

12.107x 0.5cr\177+ 1.932 qs = i0.346 cu, qs = 2.22

(qs- qt) / cu = 2.22, (4cu - qt) / cu = 2.22, qr
= 1.78 cu.

7.4 FINITE ELEMENT FORMULATION

OF THE BOUND THEOREMS

7.4.1 General re\177narks

The lower and upper bound numerical solutions are based on a finite element formulation of
the plastic limit theorems and lead to major linear programming problems. Lysmer (1970) ap-
pears to be the first to propose the use of a lower homed theorem as a solution to the linear
programming problem. A fuIl description of both methods may be found in Anderheggen &
Knopfel (1972), Bottero et al. (1980),Aysen (1987), Sloan (1989), Aysen & Sloan (1991a,
1991b and 1992). techniques assume a perfectly plastic soil model with a linear Mohr-
Both
Coulomb faiIure The solution to the lower
criterion. bound linear programming problem de-
fines a statically admissible stress field whilst the upper bound solution defines a kinematicaily
admissible velocity field. In the finite element formulation of the lower bound solution, a set
of equality and inequality constraints is constructed to satisfy the requirements of equilibrium,
stress boundary conditions and yield criterion. The collapseload is formulated in terms of un-
known stresses and maximized as an objective function to yield the corresponding collapse
load. The finite element formulation of the upper bound theorem includes equality and ine-
quality constraints to satisfy the yield criterion, the compatibility of the induced velocity dis-
continuities and the velocity boundary conditions. The objective function represents the dissi-
pated work and is minimized subject to the prescribed constraints. Both techniques bare been
applied to variety of geomechanical problems. It has been shown that in most cases the solu-
tions obtained bracket the exact collapse load within 15% or less. In some cases (e.g. plane
strain slope), especially those with high values of 0', the upper bound coIiapse load is very
much greater than the lower bound load, which makes the solution unacceptable.

7.4.2 Finite eletnent formulation of the lower bound theorem

A typical triangular Iower bound element is shown in Figure 7.29(a). Stress discontinuities oc-
cur at all edges that are shared by adjacent elemems. Thus, unlike more familiar type of finite
elements, each nodeis unique to a particular element and several nodes may share the same
coordinates. The stress components within each element vary linearly according to:

i=3 i=3 i=3

Ox = Z Nic\177xi'
=
Z
\302\242\177 Uicrzi, \"Cxz
=
Z Ni'r'xzi (7.45)
i I i=1 i 1

where Ni are linear shape functions and \177rxi, \177( and zx:i are the uodaI stresses. In order to sat-
isfy the equilibrium condition\177 the derivatives of the stress components given in Equation 7.45
are substituted into Equations 5. i. Consequently, the nodal stresses corresponding to the three

nodes of each element are subjectto two equilibrium constraints.)))

292 Soil Mechanics: BasicConcepts and Engineering Applications

\177-- X

\1772 3\".-
\177
\"-
....\177Direction
of
\"\"
\"-
\177\177extension Directions of
\342\200\242
extens[on
4

Z Z
(a) (b) (c)

Figure 7.29. Elements for a lower bound numerical analysis.

Pastor (1978) developed of the extension elements shown

the concept in Figures 7.29(b)
and 7.29(c) which enables
statically admissible stress fields to be obtained for semi-infinite
domains. For each rectangular extension element, three additional linear equalities are neces-
sary to extend the linear stress distribution to the fourth node. Each equation implies that the
stress cotnponent at the centre of each rectangleis half the sum of the stresses corresponding
to the nodes at the diagonal of the rectangle.
A typical stress discontinuity is shown in Figure 7.29(a), where the nodes in each pair of (1,
2) and (3, 4) have identical coordinates. Equilibrium of a discontinuity is enforced by ensuring
that all pairs of nodeson opposite sides of the discontinuity have equal and opposite shear and
normal stresses. This gives rise to four equality constraints. These equations are constructed by
using the stress transformation of Equations 4.1 and 4.2, which relate the normal and shear
stress on an arbitrary plane to the Cartesian stresses. If the normal and shear stresses are re-
placed by their nodal Cartesian stresses, the equality constraints are obtained in terms of the
nodal stresses.
To satisfy the boundary stress conditions, equality constraints are forced on the Cartesian

stresses for the boundary nodes. Referring to Figure 7.29(a), it is assumed that the side of the
triangular element defined by i\177odes I and 2 are located on the boundary and that the normal
and shear stresses along the boundary are known. Substituting these known values iltto the
stress transformation Equations 4.1 and 4.2, four equality constraints may be constructed.
For a lower bound solution the stresses at each node must lie within the yield surface such
that Equation 7.4 is satisfied. Since, in the numerical approach to the limit analysis, linear con-
straints are preferred, then the non-linear function expressed by Equation 7.4 is linearisedby
an inscribed polygon with p equal sides (Figure 7.30). From the geometry of Figure 7.30 it
may be shown that Equation 7.4 is equivalent to the following linearp inequalities:

t\"\177=ak(r x +bkcr z +ck\177x:

+d <-0 (7.46)

where k = 1,2, 3 .... p. For the Tresca material:

a\177:
= cos(27T.k / p), bk = -cos(27T.k / p), ck = 2 sin(2\177k / p), d = -2c u cos(\177/p) (7.47)

For a Mohr-Coulomb material:)))

 Application of Limit Analysis to Stability Probferns in Soil Mechanics 293

2\"[xz
Upper bound
I in earisatio n
\177,\177,,,,,,,/
or
\177 Mohr-Coulomb

sca material
= 2 '

K\177= .Kq ::\177

X
\177 iLin\302\260eW2r\177sba\177iUonnd

Figure 7.30. Internal and external linear approximations to the yield criterion.

a,\177
= cos(2'\177r.k / p) sinO'cos(rc/p), b\177:
= -cos(2r,..k / p) -sin \177'cos(rc/p),

c\177
= 2sin(2r\177k /
p), d = -2c'cosqb'cos(r\177/p) (7.48)

When the right-hand term of Equation 7.46 equals zero it represents the equation of the corre-
sponding side of the inscribed polygon in the stress coordinatesystem of Figure 7.30. Thus,
the inequality expressed in Equation 7.46 simply implies that the state of stressis always at the

left-hand side of the line while increasing k in the anti-clockwise direction. This erasures that
the state of stress is always inside the iinearised yield surface, thereby meeting the requirement
of the lower bound theorem. Equation 7.46 needsto be enforced at each node of each triangu-
lar element. This ensures that the linearised yield criterion is satisfied throughout each triangle
(Pastor, 1978; Aysen & Sloan, 1991a) and leads to a set of inequality constraints on the nodal

stresses. For a four-node rectangular extension element (Figure 7.29(b)) the yield criterion is
satisfied within the extension zone if the following inequalities are enforced:
F\177.I -< 0, Fk2 -< 0, Fk3 -< F\177 or Fk3
- Fkl -< 0 (7.49)

Note that the first two constraints are already included within the set of inequalities expressed
by Equation 7.46. For the triangular extension element of Figure 7.29(c) the following ine-
qualities are imposed on the nodal points:
- - < 0
F\177l -< 0, Fk2 -< Fk\177 or F\177:t Fkl -< 0, F\1773 -< Fkl or F\1773 F\177l (7.50)

The first constraint in Equation 7.50 may be included in the assembly of inequalities defined
in Equation 7.46, while the two remaining have the same form of constraints described by
Equation 7.49. It is now required to define the objective function Q that represents the collapse
load applied over a specified length of the boundary (as an example); this is given by:
Q =/(\177rnl +on2)]2 (7.51))))
294 Sol[ Mechanics: BasicConcepts and Engineering Applications

Figure 7.3 I. Elements for an upper bound numerical analysis.

where I is the length of tim edge and on2

\177,\1771, are the unknown normal stresses at its two ends.
Using the transformation
stress equations, objective the function can be expressed in terms of
the nodal stresses at its two ends. The elemental constraints and objective function now are as-
sembled by the usual rules of finite elements to give the overall linear programming problem.

7,4.3 Fi\177ite element formulation of the upper bound theoretn

Figure 7.31 shows the triangular elements that are used to model the velocity field under plane
strain loading, Each node has two velocities and each element has p plastic multiplier rates
(Equation 7. i) where p is the number of sides in the line\177ised yield polygon, The velocities
vary linearly throughout each element according to:

i=3 i=3
u = \177Niu i, v = \177Niv i (7.52)
i=1 i 1

where linear shape functions

Ni are and ui, vi are the nodal velocities in the x and z directions
respectively. Whilst the linew triangular element, or the constant strain triangle, is considered
to be an unsophisticated element in the general finite element formulation, in the numerical
application of the upper bound theorem it has been used extensively to solve quite difficult
problems. The shape functions Ni can be Ibrmulated in terms of global coordinates and the
co\177esponding equations can be found in many texts (e.g. Zienkiewicz, 1970)and will not \177
discussed hem. Combining Equation 7.1 (plane strain deformation of a rigid-peflEctly plastic
soil) and Equation 5.3 (general definition of a plane strain deformation), we obtain:

g{ =
- ' = = = + = eL (7.53)
\177f:
aN -g \177' g

where F represents the yield criterion and dL is a plastic

multipfier; compressive strains are

taken as positive, in oNer to Equation 7.53 and thereby provide

remove the stress te\177s from
a linear relationship between the unknown velocities and plastic multiplier rates, an external
approximation to the yield criterion is needed. A six-sided approximation is shown in Figure

7.30. The linearised yield condition may be written as:)))

 Application of Limit Analysis to Stability Problems in Soil Mechanics 295

Fk -ak(\177 x +bkcy z +Ck'\177xz

+d = 0 (7.54)

where k = 1, 2, 3 .... p. For a Tresca material:

ak =
cos(2\177k/p), b k = -cos(2rd:/p), ck =
2sin2r\177k/p), d = -2c u (7.55)

For a Mohr-Coulomb material:

a\177:
= cos(2rr2 / p) - sin \177', b\177:
= cos(2\177k / p)
- sin 0',

c\177
=
2sin(2\177k/p), d = 2c'cosq\302\242 (7.56)

Differentiating Equations 7.52 and 7,54 and substituting into Equation 7,53 furnishes three
flow rule equations in which the stress terms are removed.

i=3\177N ' k=p i=3\177N

k=l = k=l

i 30N i=30N
\"
,,. +y\177
' ',\177+ y\177dZ.\177
=0 (7.57)
ax i=1 oz k=|
The geometry of a velocity discontinuity element aud the configuration of the nodes can be
describedsimilar to the stress discontinuity and is shown in Figure 7.3 I. For each velocity dis-
continuity typestwo are necessary to enforce
of constraints the following conditions: (a) the
work dissipated by relative sliding along the discontinuity (imernal work) is positive which
generates a set of inequalities to ensure the sigu convention for the discontinuity, (b) the flow
rule must be satisfied within the discontinuity for the plastic deformation. For a Tresca mate-
in the normal u,\177 0 (zero dilation). For a Mohr-Cou[omb
rial the jump = material the
velocity
flow rule in the discontinuity is satisfied if an equality constraint is enforced to ensure that the
angle between the discontinuity and the direction of the velocity jump is equal to \177'. The final
type of equality constraints to be imposed on the nodal velocities arise from the boundary con-
ditions. The objectivefunction is the linear expression for the dissipated work, which has two
terms. Its first term expresses the sum of the work dissipated within the velocity discontinui-
ties. In a Tresca material and for one velocity disco,\177tinuity, this term is:

where I is the length of a discontinuity and lull is the absolute value of the tangential velocity
jump. This equation is integrated to express E\177I in terms of the nodal velocities. The second
term expresses the work dissipated by the plastic stresses for triangular elements. In a Tresca
material and for one triangle, it can be shown that this te\177 is:

k=p
= + + = (7.59)
Ei2 \177(oxg.r oz\177 z zxz\177a.:)\177 2cuA \177dL t
A k-1)))
296 Soil Mechanics: Basic Concepts and Engineering Applications

10

9
Upper bound: FIgM \177

Lower bound: FEM

I r i | I i I I

1 2 3 4 5 6 7 8 9 10

Figure 7.32. Lower and upper bounds for a tunnel heading compared with FDM. yB / c, = O.

where A is the area of the triangular element. For a Mohr-Coulomb material, an equation simi-
lar to Equation 7.59 can be obtained. Once the various coefficients are assembled the dissi-

pated work is minimized and a kinemadcally admissible velocity field can be found. The stress
components at each node are obtained by substituting the calculated velocities and plastic mul-

tipliers in the set of Equations 7.53.

We note in passing that the set of equality and inequality constraints may be solved very ef-
ficiently by applying the steepest edge active set algorithm, which fully exploits the extreme
sparsity of the constraints. As a result, very large analyses may be conducted on a workstation
or microcomputer.

7.4.4 Applications to slopes and tunnels

A sum_mary of stability numerical methods for slopes are

bounds obtained from the above two
shown in Tables 7.1(a) and 7.lib).
and square tunnels,
Applications and the tunnel for circular
heading, are shown in Tables 7.3, 7.4 and 7.5 respectively. For the plane strain heading a finite
difference method developed by Cundall (1980 and 1987) was used by the author to verify the

effectiveness of the bound solutions. The results of these comparisonsare shown in Table 7.5
and Figure 4.32.
For the active failure of the tunnel heading (where the heading collapses towards inside the
tunnel) and for 7B / c, > 0, it can be shown that the upper bound value of the load parameter
(qs - q\177)/ cu is:

(qs -qt)/cu
= [(qs -qt )/c. =0
-
(gB/c. )(H / B + 1/2) (7,60)
]yg/,.,,

The following formula for undrained conditions is suggested by the author:

- c,, = 4.0 + 2 l n(H - B + 1/ 2)

(qs qt )/ / B) (yB /
c. )(H / (7.61))))
 Application of Limit Analysis to Stability Problems in Soil Mechanics 297

Whilst the results from this equation may be slightly higher or lower than the exact solution,
they are always bounded by the lower and upper bounds with reasonable accuracy.

Table7.3.Undrained stability of a shallow circular tunnel (Aysen & Sloan, 1991b).

yD / c, = O \177lD/ cu = l \302\245D/c,=2 \177lD/ cu = 3

HID
LB UB LB UB LB UB LB UB
\177 d - 1.60d
1 2,27 2.55`/
`/
!.08 \17793 -0.16 -0.16 -1.60
- 1.35
d
- 3.77d
2 3.25 3.48 0.97 -1.4o 3.87
a
- d -
3 3.78 4.29 &47 0.88`/ 2.95 - 2.53 6,49 5.96`/
'\177
4 4.30 5+03`/
d
-0.08 0,62`/ -4.57 -3.80`/ 9.11 -893
0.74 0.28`/ -
d
- d
5 4.65 5.50 6,20 -5.14 !1.80 1\177)\17757
d
6 4.48 5.90 -1.40 -0.35 -7.90 -6,65 14.50 -13,10`/
7 5.30 6.30 -2,10-1.05 -9.65 -8.40 -17.48 -15.75
8 5.48 6.65 2.95 -1.75 - 11.50-10.15-20.40 -18.55
9 5.75 6.90 -3.80 -2.50 -13.46-11,90 -23.17 -21.30
I0 5.92 7.15 -4.71 -3.22 -15.40-13,60 -26.10 -24.06
Note: The numbers with superscripts d indicate a collapse mechanism
(improved version of Figure 7.24(d)) which gives best upper bound.

Table 7+4. Undrained stability of a shallow square tunnel (Aysen & Sloan, 1991a).

yB / c\177= O =
\"tB / c\177 l / c, = 2
\302\245B / c\177
\302\245B
= 3
H/B
LB UB LB UB LB UB LB UB

a a a a
0.3 0.522 0.600 0.261 0.300 -0.050 0.000 -0.401 -Q300
a a
- 0+055 a - 0.590 - a
0.4 0.712 0.800 0.344 0.400\177 0.000 0.400
0.5 0.915 ].00
a
0.477 0.500 - 0.072 0.000\" - 0.760 - 0.500 \177
0.985 -
\177'
0.6 1.12 1.20\" 0.538 0.600'\177 - 0.124 0.000\" 0,656\177'
'\177 '\177 \177'
0.7 1.32 1.40 0.631 0,700 -0.181 0.000 1.21 - 0.874
0.8 1.52 1+60\" 0.704 0.800
'\177
-0.261 0.000\" 1.41 - t.102'
a b b
0.9 1.70 1.80\" 0.758 0.900 - 0.343 -0.115 1.62 - 1.32
# a - 0.440 z'
- 1.83 - 1.54z'
1.0 1.88 2.00 0.810 1.00 0,234

1.5 2.50 2,77

#
0.750 0.960ff - 1.09 - 0.854
t'
301 2.682'
b t' b z'
2\1770 2.89 3.15
v
0.600 0,825
t'
1.78 1.50\177 4 \"\177v 3.84
t'
3.0 3,45 3.82 0,140 0.478 3.24 2,87 61\"\177 6.23
t'
4.0 3.90 4.37 - 0.410 0+039 4.83 4.31
\177
9.38 8.68
5.0 4.23 4.79 - 1,06 -0.554 -6.51 -5.88 -12.10 -11.21
6.0 4.55 5,19 -1.78 -1.15 -8.25 -7.48 -14.87-13.81
7.0 4.81 5.57 -2.57 - 1,77 - 10,05 -9.10 - 17,65 - 16.43

8.0 5.03 5.79 - 3.36 -2.54 - 11.87 - 10.87 - 20+50 - 19.21

9.0 5.28 6.05 - 4.10 - 3.29 - 13,79 - 12.62 - 23.57 - 21+95
10.0 5.47 6.29 -5,00 -4.05 -15.79 - 14+38 -26.60 -24.71

Note: The numbers with superscripts a, b (corresponding to Figures 7.25(a) and

7.25(b))indicate the mechanism which gives best upper bound.)))
298 Soil Mechanics: Basic Concepts and Engh, eering Applications

Table 7.5. Undrained stabifity of a plane strain heading (Aysen & Sloan, 199lc).
yB / cu =0 = [
/ cl\177
\177{B yB / cu
= 2 yB / cu
= 3
H/B
LB FDM UB LB UB LB UB LB UB

1.0 4.00 4.05 4.39 2.46 2,89 0.85 1.39 - 0,74 0.11
[.5 4,60 4.97 5.[3 2-56 3.13 0.49 [.13 -1.76 -0.87
2\303\2670 5.05 5.57 5.68 2.40 3,18 - 0.20 0.68 - 2,8\177, - 1.82
3.0 5,75 6.35 6.50 2.20 3.00 - 1,40 -0.50 - 5.03 - 4.00
4.0 6.25 6.92 7,21 1,7[ 2.71 -2.86 -I.79 -7.49 -6.29
5.0 6.70 7.37 7,70 1.15 2.20 -4.48 -3.30 -10,11 -8.80
6.0 7.02 7.70 8,12 OAt 1.62 -6.10 -4.88 - 12.90 - 11.38
7,0 7.33 8.03 8.49 - 0.32 0.00 - 7.90 - 6.51 - 15,53 - 14.01
8,0 7.46 8.28 8,83 - 1.21 0.33 -9.80 -8.17 - 18.43 -16.67
9.0 7.57 8.53 9.08 -2.16 -0,42 -11.75 -9,92 -21.39 -19,42
10,0 7.70 8.71 9.32 -3.16 -I.18 -13.75 -11.68 -24.39 -22.18

7.5 LIMIT EQUILIBRIUM METHOD AND CONCLUDING REMARKS

7.5.1 The concept of the limit equilibrium method

The theory of limit equilibrium uses the combined features of the bound theorems and calcu-
lates the collapse loads that are neither lower bound nor upper bound. Its application to soil
mechanics problems,particularly in the areas of lateral earth pressures, stability of slopes and
bearing capacity problems
- has beenvery successful (Chapters 8 to 10). In a limit equilibrium

method a collapse mechanism is selected to represent the failure of the material. Each block of
the mechanism is assumed to be a rigid-perfectly plastic material, It is assumed that on the slip
or sliding surfaces the mobilized shear stress zm, satisfies the following equation:

F = \"r2f [T,
m (7,62)

where zf is the shear the soil and F is the factor of safety. The factor of safety is as-
strength of
sumed constant along surfaces. If the loading
all sliding is increased and the shear stresses on
the sliding surfaces are forced to approach \"cj-(defined by an appropriate failure criterion), then
the material will fail by following the pattern of the collapse mechanism, The factor of safety
for an assumed mechanism is calculated by applying the equilibrium conditions to the mecha-
nism. An iterative procedure is carriedout to establish the optimised values of the variables of
the mechanism and subsequently the factor of safety. For F = 1 the collapse mechanism will
yield a collapse load that is usually between the lower and upper bounds.
If the number of unknown forces is larger than the number of available equilibrium equa-
tions, then some simplifying (but reasonable) assumptions have to be made.Using the equilib-

rium equations, the collapse load, or any stability related parameter for F -> 1, may be formu-
lated in terms of the shear strength characteristics for an optimised mechanism. If the
formulation is not possible, then a trial and error method has to be employed to achieve the op-

timal mechanism.)))
 Application of Limit Analysis to Stabiti\177' Problems in Soil Mechanics 299

7.5.2 Concluding remarks

The lower and upper bound theorems of classical plasticity provide powerful methods for the
evaluation of various stability problems in geotechnical engineering. The efficiency of the
both techniques can be verified by careful analysis using numerical methods (e.g. displace-
ment type finite elements and explicit finite difference method).
In a lower bound solution a statically admissible stress field is constructedso that it satisfies
the equilibrium, stress boundary conditions and the yield criterion. The rigid-perfectly plastic
material is assumed to obey either the Tresca or Mohr-Coulomb yield criteria. Points repre-
senting the state of stress are Iocatedinside the convex yield surface and therefore the material
does not fail.
In an upper bound solution, a kinematically admissible velocity field is constructed so that
it satisfies the compatibility, velocity boundary conditions and yield criteria but not necessarily
equilibrium. The points representing the state of stress are located on the yield surface and
thus the material fails under the computed collapse load.
In both methods manual cmnputations can be used to estimate the corresponding collapse
load. The exact solution lies between the two bounds. In the lower bound solution the selection
of the stress field surrounded by the stress discontinuities needs ingenuity and practice. In the
upper bound solution a convenient collapse mechanism yields results that are reasonably close
to the exact solution. Finite element formulation of both \177nethods leads to a large linear pro-
gramming problem and needs a suitable algorithm to obtain a feasible solution.

7.6 PROBLEMS

7.1 A stress discontinuity makes an angle

\302\260
of 60 with the x-axis. On the right-hand side of
the \"rx: = 50 kPa. Determine:
=
c\177x 100 \"kPa,
discontinuity:

(a) the magnitude of \177: to satisfy the failure criterion.

(b) the normal and shear stresses on the discontinuity, and
(c) the state of the stresses at points on the left-hand side of the stress discontinuity to

satisfy the failure criterion.

-
Answers: % (at R) = 327.2 kPa. \177 200.1 \"kPa, z- 123.4 \"kPa, c\177:(at L) =521.8 kPa, \177x

(at L) = 164.8 kPa, Zx.: (at L) =

- 62.3 kPa
7.2 Resolve Example 7.3 by taking into account the weight of the material. Assu\177ne the ver-
tical stress at every point increases by)z, unit
(\302\245= weight, z = depth).
Answer: 5cl\177

7.3 Calculate the bearing capacity factor N,. for undrained conditions for a smooth strip foot-
\302\260
ing if the ground surface outside of the foundation \177nakes an angle of 15 above or below
the horizontal foundation level.

A\177lswers: 5.17, 4.43

7.4 For the plane strain slope sbo\177vn in Figure 7.22(a). calculate the lower and upper bound
values for q / c,. For the lower bound solution select a reasonable value for the number)))
300 Soil Mecl\177anics: Basic Concepts and Engineering Applications

of the stress discontinuities passing through point d and apply the concept used in Equa-
tion 7.11. For the upper bound solution use the mechanism shown in Figure 7.22(a). As-
sume the stability number 7H/cu = 0.
Answers: 3.57 (lower bound), 4.10 (upper bound)
7.5 A plane strain vertical cut is subjected to a vertical uniform load q at the upper ground
surface. If the lower bound for the stability number'yH/c' (with q/c'
= 0) is 6.12, (this
is obtained from a finite element analysis) calculatea lower bound for the load parameter
q / c' for the given soil parameters.
c\" = 20 kPa, =
7 = \177,9kN/m 3,
t\177' 30 \302\260, and H = 5 m.
Answer: 2.7
\302\260
7.6 For a 45 plane strain slope in a c,,, t\177u= 0 soil, compute the upper bound values for the

stability number yH/cu using the collapse mechanism shown in Figure 7.18(a) for:
(a) a single block with or1 = 22.5% and
(b) two blocks with [3t = 75 \177,cq= 45 \302\260, =
c\1772 15 \177.

Answers: 9.67, 7.77

\302\260
7.7 For a 45 plane strain slope in a cu, 9\177= 0 soit. calculate the load parameter q / ct\177using

the collapse mechanism of Figure 7.18(a) together with the following data:

yH / c, = 0, =
[\177t 60 \302\260,
\177
= 60 \177,c\1772= 10 \302\260.

Answer: 3.88

7.8 A 10 m height of saturated clay is supported by a rough retaining wall. The propertiesof
the so\177l are: c, = 60 kPa, t\177u= 0, and y = 18 kN/m. The vertical boundary load as q = 50
kPa and cw = 30 kPa. Calculate the lower bound value of the horizontal active thrust and
the position of its point of application.

Answers: 210.0 kN, 1.61 m

7.9 A rough wall of height 6 m retains a sandy soil for which c' = 0 and 9' = 35\302\260,
retainin\177
19 kN/m\". A uniform boundary
and \302\245= load ofq = 50 kPa acts on the upper ground sur-
face. Assuming \177i'= 35 \302\260,
calculate the lower and upper bounds for the active thrust (the
resuItant of the horizontal and vertical components) on the wall. For the upper bound
solution assume a single variable mechanism similar to Figure 7.23(a).
Answers: 176.9 kN (lower bouud), 160.3 kN (upper bound)

7.10 Using the single variable mechanism of Figure 7.23(a), find the upper bound value for
tl\177e active using the following
thrust data:
H = 5 m, ct = 34 \302\260,
q
= 80 kPa, 6' = 20 c' = 10 kPa, 9' = 25\302\260,
\302\260, and y = 18 kN/m 3.

Answer: 165.5 kN
7.11 For the upper bound mechanism of Figure 7.26(a) compute an upper bound value for qt.
The given data are:

o\177=[3=60\302\260,6=90 H=lOm,
\302\260, D=2m, cu =40kPa, gu=0, y=20kN/m 3.

Answer: 399.2 kPa (tensile))))

 Application of Limit Analysis to Stabilit), Problems in Soil Mechanics 301

7.7 REFERENCES

Anderheggen, H. 1972. Finite

E. & Knopfel, element limit analysis using linear programming. Intern.
journal of solids and (8): 1413-143.
structures
Alkinson, J.H. 1993.An introduction to the mechanics of soils and foundations. London: McGraw-Hill.
Aysen, A. 1987. Lower bound solution for soil mechanics problems using finite element method, Proc.
2 \177at\177onal cor\177f o\177sod mechantcs andfou\177rdation engineeringl: 121-136. University of Bogazici,
lstanbul, Turkey.
Aysen, A. & Loadwick, P. 1995. Stability of slopes in cohesive frictional soil using upper bound collapse
\302\242\177
mechanisms and numerical methods. Proc. 14 Australasian conf on the mechanics of structures and
materials 1: 55-59.
Aysen, A. & Sloan, S.W. 1991a. Undrained stability of shallow square tunnel. Journal GE, ASCE,
117(8): 1152-1173. \177h
Aysen, & SIoan, S. W. 1991b. Stabiiity
A\303\267 of a circular tunnel in a cohesive frictional soil. Proc. 6 in-
tern. coal in Australia on finite element methods. University of Sydney. 1: 68-76.
Aysen, A. & Sloan, S.W. 1991c. Undrained stability of a plane strain heading. Researcfi Report No.
059.02.1991, lSBN 0 725907134.NSW, Australia: The University of Newcastle.
\177tr
Aysen, A. & Sloan, S.W. 1992.Stability of slopes in cohesive frictional soil. Proc. 6 Australia-New
Zealand conf on geomechanics:Geotechnical risk-identification, evaluation and solutions: 414-419.
New Zealand: New Zealand Geomechanics Society.
Bottero, A., Negre,R., Pastor, J. & Turgeman, S. 1980. Finite element method and limit analysis: Theory
for soil mechan Computer methods in applied mechanics and engineering (22): 131-149.
ics problems.
Britto, A.M. & Kusakabe, O. 1985,Upper bound mechanisms for undrained axisymmetric problems.
Proc. 5\177'intern, conf on mcmerical methods in geomechanics. Nagoya. 1691-1698.
Carter, J,P,, & Balaam, N.P, 1990. Program AFENA, a general finite element algorithm. Centre for
tech\177tical researeh, University of Sydney, Australia.
Chert, W.F. 1975.Li\177nit a\177alysis and soil Plasticity, Amsterdam: Elsevier.
Chen, W.F. & Baladi, G.Y. 1985, Soil plasticity: theory, and implementation. Amsterdam: Elsevier.
Chen\177 W.F. & Saleeb, A.F. 1982.Constitutive equations for engineering materials, Vol. I- Elasticity and
modeling. New York: Wiley-lmerscience.
Chert, W.F. & Saleeb,A.F. 1986. Constitutive equations for er\177gineering materials, Vol.2- Pfasticio' and
modelling. New York: Wiley-lnterscience.
Cundall. distinct element program
1980. UDEC - A generalized
P\303\267A. for modelling jointed rock. Peter
Cundall Report PCAR-I-80, US Army, European
Associates. Research Office.
Cundall, P.A, 1987.Distinct element models of rock and soil structure. \177nE. T. Brown (ed), Analytical

and computational methods in engineering rock mechanics: 129-139. London: George Allen & Un-

Davis, E.H., Gunn, M.J.. Malt, R.J. & Senevirame, H.N. 1980. The stability of shallow tunnels and un-
derground openings in cohesive matenal\177 Geotechnique 30(4): 397-4[6.
Heyman, J. 1973. The stability of a vertical cut. Intern. jourm\177l of mechanical science (15):845-85\177.
Lysmer, J. 1970. Limit analysis of plane problems in soil mechanics. Journal GE, ASCE 96(4):1311-
1334.
Mair, R.J. 1979. Centrifugal modelling of tunnel construction in soft clay. Ph.D. Thesis, Cambridge Uni-
versity,
Mulhaus, H.B. 1985.Lower bound solutions for circular \177unnels in two \177t\177d
\177hree dimensions. Journal of
rock mechanics a\177d rock et\177gineering 18: 37-52.)))
302 Soil Mechanics: Basic Concepts and Engineering Applications

Naylor, D..I. & Pande G,N. 198I. Fir\177ite elements in geotechtffcal engineering. Swansea: Pineridge Press.
Parry, R.H.G, 1995.Mohr stress paths and geotechnics. London:
circles, E & Spon.
Pastor, J. 1978.Limit analys\177s: numerical de\177.ern:tination of compIete slatical solutions: application \177o\177.lle

vertical cut, Journal de mecanique appliquee, (in French) (2): [67-196.

Sloan, S.W. I988. Lower bound limit analysts using finite elements and linear progrwnming. Intern+
journal for mrmerieal and
analytical methods in geomechanics 12(!): 61-77.
$[oan, S.W. 1989.Upper bound limit analysis using finite elements and linear Programming. fntern.
journal for numerical and analytical methods in geomechanics 13(3): 263-282.
Sloan, S.W. & Aysen, A. 1992.Stability of shallow tunnels in soft ground, in G,T. Houlsby & A.N.
Schofie\177d (eds), Predictive soil mechanics. London: Thomas Telford,
Turgeman, S. & Pastor, J. 1982. Limit analys\177s: a linear formulation of the kinematic approach for axi-
symmetric mechanic problems. Intern. journal for numerical and analytical methods in geomechanics
6: 109-128.
Zienkiewicz, O.C. 1970. The finite element method in engineeting science. London: McGraw-Hill.)))
CHAPTER 8

Lateral Earth Pressure and Retaining Walls

8.1 INTRODUCTION

A retaining wall is a \177truclure thai

supports a vertical or ucar vertical IScc of soil and is used
facilitatc a maj(\177 change ground sulkace level.Thc resulting
in the horizontal stress from the
\177oil on the wall is cmnmonly referred to as lateral eaclt\177 pressure. In tile evaluation of the
magnitude and distribution of th\177s lateral earth pressure, it is assumed that the soil behind the
wall (backfill) is \177mthe verge t\177ffail ure and obeys either the Tresca or Mohr-Coulomb fmture
crileri\177)n. Failure of the backfill \177ccurs by two mechanisms depending \177mtile directitm of the
wall displacement, If lhe disptacecnent {)f the wall is away \177he hackfill
t\302\245t)m soil the resultiug
failure is called acgh'\177,. A pa.\177.,i\177'e failure occurs if the wall is displaced l(ywards the hackfill
until the limiting di\177placcmcn\177 is achieved. When the wall has n() d\177spIacement, lateraI
\177t\177e

earth pressmc a\177-\177c,.v\177 can he estimated from elastic equJlJhritt\177ll or field measurements, hi

general lhcrc are basic thc{\177ms used \177o estimate Interal earth pressures, viz., the Rankine
me\177hod and the Terzaghi and Peck me\177hod. Rankine's theory assumes a long and
vertical wall where \177hc Imeral p,essure increases linearly with depth. This zmalysis, although
conve\177ient, imposes \177mc Ihnitatioas and conse\177iuentiaI e\177ors, I4ur a \177'all wilt\177 sign[/]cant
friction, or il: lhere is \177mirregular grtmnd \177urface. a Coult\177mb wedge analysis, i\177ctmjtmclion

with the charts im\177duced by Tcrzaghi & Peck (1967}. is appIied.In this method the actuaI
lateral pressure dist\177 ibuti(m \177stlOldelined but, based o\177 cxperimemal data. the locatit\177l of thc
resullantof these \177lateral) pressures \177sdefi ned. Ht)wever, fl\177r lung i] ictitmal walls such as sheet

piles, Rankine's mcth\177)d is applicd using the earth pressure c(>ci]tcienls t\177btained from a
Coulmnl\177 wedge analysts. This c(m\177bination, although the(\177ret}caIly inc\177n'ect, api\177em\177 \177ogive
rcasonaMc eslimale5 the
\177)\177 magnitude and d\177stnhutit\177n of the lateral s\177)il pressure.

8.2 EARTI[ t'RESSIJRE AT-REST

The m\177gnitude the

{\1771 lalcral l\177lessu\177e depends on lhc type \177l\"the wall, II\177eamount ()f the wall

displacemc\177l, the type of (backfill) s()iI, lhc drainage cond\177tion\177 behind the walt, and the mag-
303)))
304 Soil Mechanics: Basic Concepts and Engineering Applications

nitude of any externally applied surface loading. tn the absence of any lateral expansion or
compression of the soil, the state of stress within the soiI is termed as at-rest, In general the
two-dimensional stress fieId immediately behind a retaining wall is indeterminate and only the
effective vertical stress can be defined in simpIe cases. In a homogeneous half-spacethe effec-
tive vertical stress at depth z is defined by Equation 2. i:

(itz

where y is the unit weight and u is the pore pressure at depth z. For a homogeneous, isotropic
and elastic soil with zero lateral strain the lateral stress O\"x (Figure 8. i)is expressed by:
, = kocr[_
(l x = O\" (8.1)
1-\177

where bt is Poisson's ratio whose

or field deternaination is problematic, and ko is the
Iaboratory
coefficient of soil pressure condition. Equation
for the at-rest
8.1 implies that the axial strains
ex, and ey (y is normal to the plane of the page) are both zero. The magnitude of ko varies from
0.42 to 0.66 depending on the type of the soil, its degree of compaction and plasticity charac-

teristics, and the degree of its disturbance (Bishop, i958). For truly normally consolidated
soils that exhibit zero cohesion during drained shear, a value for ko may be calculated from the
following generally accepted empiricaI equation (J:iky, 1948):

= 1
k,\177 -sin0' (8.2)

For overconsolidated soils the vaIue of ko is higher than that given by Equation 8.2. Alpan

(1967) suggested the following relationship:

k\302\260'\302\260ce \177
=
(OCR) (8.3)
ko,NC
where ko,oc R and ko,NC are the coefficients of earth pressure at-rest for overconsolidated and
normally consolidated soil respectiveIy, OCR is the overconsolidation ratio and n is a number
depending on the plasticity characteristics of the soiI. Wroth & Houlsby (i985) give tt = 0.42
for soil with low plasticity (P1 < 40%) and n = 0.32 for soil with high plasticity (PI > 40%).

Z
Backfill soil
_?a, z
c', e\" or c u,
Cu

Figure 8.1. S\177ate of stress behind a re\177aining wall.)))

 Lateral Earth Pressure and Reta#ffng Walls 305

After a statistical analysis of reported data, Mayne & Kulhawy (1982) proposed that n = sin\177'.

Hence,

(1 -
sin *'
ko = sin \177')(OCR) (8.4)
Most laboratory methods for the determination of lateral earth pressures are based on triaxial
and odeometer tests. A full discussion on the laboratory measurement of lateral stress in soft
clays may be found in Benoit & Lutenegger (1993).In-situ measurements used to determine
ko involve the use of SPT and CPT type field tests that give rise to empirical correlations but,
unfortunately, them is usually a wide scatter in the actual data, especially for soft clays (Kul-
hawy et al., 1989).

8.2.2 Earth pressure due to surface loading and compaction

The additional stress along a wall caused by surface
lateral loading can be estimated from a

Boussinesq type distribution. In a smooth wall the lateral component of stress at-rest is dou-
bled to take account the presence of the rigid smooth wall. This doubling is equivalent to the
application of a mirror of the real load (symmetric about the back face of the wall) that en-
forces zero lateral strain and zero shear stress on the back face of the wall.
The increaseof lateral earth pressure due to the effect of compaction equipment moving
across the backfill has been investigated thoroughly and there is (field) evidence that shows a
considerable increase in lateral pressures on the wall. This increase, which is regarded as a re-
sidual lateral pressure, is concentrated in the first few metres. The state of stressapproaches
the ko state at a depth of approximately 9 m (Duncan & Seed, 1986).Following these findings,
Williams et al. (1987)developed charts and tables relating the induced lateral stress to the type
of the compaction equipment.

8.3 RANKINE'S THEORY FOR ACTIVE AND PASSIVE SOILPRESSURES

8.3.1 The concepts of active and passive states and wall displacetnent

Rankine's theory of active and passive requires a long smooth

soil pressures wall and assumes
a linear distribution earth pressure. Theseassumptions
of lateral were challenged by Teczaghi
(during 1930s) when he developed a Coulomb wedge analysis. This was a turning point in the

computation of the resultant of the lateral earth pressure and the determination of its point of
application behind a frictional wall (Peck, 1990). Today, the Coulomb wedge analysis has
changed from its original computational-graphical method to a computer-based method.
Rankine's theory, along with the lateral earth pressure coefficients obtained from the Coulomb
wedge analysis and with numerical modelling of the retaining system, are applied to long re-
taining structures such as sheet piles.
Considerthe retaining wall shown in Figure 8.1 to be subjected to a horizontal displace-
merit away from the retained soil. This displacementcausesan expansive strain within the re-
tained soil where the affected volume of the soil depends on the height of the retaining wall. If)))
306 Soil Mechanics: Basic Concepts and Engineering Applications

the height of the wall is infinitely long then the affected voIulne will progress to infinity to the
right of the retaining wall. However, for a finite height of wall, the finite volume of the soil
will be subjected to horizontal expansive strains. As a result, the lateral earth pressure acting
on the wall will decrease and eventually the soil behind the wall will fail. This means that the
state of the stress within the soil is transferred from elastic equilibrium to plastic equilibrium.
If there is no surface loading then the magnitude of the effective stress (defined by:
vertical
o\177z
= 7z - u) remains constant. In the presence of a surface loading q, the effective vertical
stress increasesin a Boussinesq pattern. At failure, the vertical contribution of the surface
loading at every point becomes equal to q. During the progress of the lateral displacement, the
diameter of the Mohr's circle of stress increases until at failure, it becomes tangent to the fail-
ure envelope. The lateral stress, which is the lateral soil pressure, becomes a minimum at
every point along the wall and it is termed as active earth (or soil) pressure as it refers to the
active state of failure behind the retaining wail. For a long smooth wall, the major principal
stress is vertical (\177t = \"\177z- u). The minor principal stress represents the active soil pressure
and acts horizontally. The magnitude of this pressure can be found from the relationship estab-
lis\177ed between cr'l and 0\"3 at failure (Equation 4.13 or 4.14).
If the wall moves towards the backfill the lateral pressure will increase, reaching its maxi-
mum value called the passive earth (or soil) pressure. This representsa passive failure behind
the wall for which the effective vertical stress represents the minor principal stress
(o\1773
=
7z
- u), and the i\177orizontal major principal stress representsthe passive earth pressure.
The magnitude of the pressure at any depth can be found by substituting 0'3 into Equations
4.13 or 4.14. The lateral pressure-displacement ratio (A / H) behaviour is shown in Figure 8.2
from which it can be seen that to achieve a passive failure, more displacementof the wall (in
comparison with the active failure) is needed. Duncal\177 et al. (1990) summarized the existing
laboratory tests on model retaining walls and controlled field experiments that have been car-
ried out from 1934 to 1990. According to d\177is report, the ratio ,\177/H required to reduce earth
pressure to active values in sands, silty sands and sandy gravels, varies from 0.0003 to 0.008
for rotation and from 0.001 to 0.005 for translation, wid\177 d\177e larger values of A / H applicable
to larger walls. In the reported data two definitions of active failure have been used. T\177e first
corresponds to the active thrust required to reach its minimum value whilst the second defini-
tion corresponds to the development of a linear lateral eart\177 pressure. In many experiments the

+A

- At rest pressure

- Pa Active pressure
\177--

Figure 8.2. Displacement requirements for active and passive sta\177es of failure.)))
 Lateral Earth Pressure and Retainhrg Wails 307

first definition is achieved while having a non-linear earth pressure at an elevation higher than
0.33H, Conduuing displacement results in a linear distribution of pressure and the resultant at
0.33H. Displacement ratios required are between 0.005 to 0.132 for rotation (higher values as
the clay percentageincreases)and an average value of 0.025 for translation, Changes in the
lateral earth pressure with time have been reportedby number of investigators. Field observa-
tions have shown that the earth pressure changes with time and approaches the at-rest values.
For a long-term active state, the average of the active and at-rest conditions may be used in de-
sign (Munfakh, 1990).

8.3.2 Linearearth pressure distribution for a smooth wall and horizontal ground surface

Figures 8,3(a) and 8.3(b) provide illustrations of the and passive failure conditions re-
active

spectively whilst the states of stress are shown in Figure 8.3(c), The inclinations of the poten-
tial failure planes from the horizontal grou\177d surface are 45\" + \177' / 2 and 45 \302\260
-
\177' / 2 for the

active and passive states respectively. For the active state, the lateral earth pressure at depth z
is \1773 =pL\177 and \177]
=
#:=yz
- u. these into Equation 4,13 we obtain:
Substituting

Pa
= \177, 1-sin\177b'
2c\177
sinqb'
->
Pa
=
ci-ka (8.5)
,/\177-

(a)

(c)
Figure 8.3. Active and passive stales of stress at the back of a long smooth retair\177ing wall,)))
308 Soil Mechanics: Basic Concepts and Engineering Appficatfons

where ka is the lateral active earth pressure coefficient given by:

k a = (1 - sin qb')/(1+ sin 0') = tan:Z(45\302\260-\177'/2) (8.6)

If a uniformly distributed vertical load q is applied at the ground surface of the backfill, then:

= -
Pa (cr'z
+
q)ka 2c'\177a (8.7)
The general form of the lateral earth pressure for the passive state is:
Pp
= (0'z + + (8.8)
q)kp 2c'k\177v
where kp is the lateral passive earth pressure coefficient given by:
= (i + sin - sin
(b') = tan
2
(45
\302\260
+ \177'/2) (8.9)
kp \302\242')/(1

Equations 8.7 and 8.8 quantify the Iinear increase in Iateral earth with increasing
pressure
depth z. Assuming that the same distribution applies for a wall of then, for a ho-
finite height,
mogeneous soil with no water in the backfill, o\"z
=
yz. The total active or passive thrust is the
integral of the lateral earth pressure behind the wall.

= 0.5yH 2k + qHk - 2c'H

Pa a a \177a (8.10)

= \177-2c'H (8.11)
Pp 0,5yH2/\177p +qHkp k\177v

Note that for a purely cohesive saturated clay with undrained shear strength parameters of cu
and 0u = 0, then ka = kp = I. Figure 8,4 simws the distribution of lateral active earth pressure
behind a smooth wail. Figure 8.4(a) represents the case where there is no surface loading. At z
= 0 the soil behind the wall is to a maximum stress of 2c'\177ko, The tensile
subjected
tensile
stress decreases with becomes zero at depth
depth and Zo. Since it is common to ignore the
tensile strength of the soil, the cohesion is the minimum shear strength at zero normal stress
and the failure criterion does not apply for the tensile normal stress. Consequently cracks will
develop down to the depth zo, Figure 8.4(b) represents the effect of surface loading in the form
of uniform lateral pressure equal to qka. The combined effects of gravity and surface loading
are shown in Figure 8.4(c). Notc that the area between the Iine representing the earth pressure
and the vertical axis is equa[ to the total active thrust. If there is no hydrostatic water pressure,
vertical =
stress corresponding to zo is c\302\242:
the effective Yzo, Upon substituting this into Equa-
tion 8.7 and setting the active pressure to zero we obtain:

zo = 2c'/(\302\245/@\177-\177-
a ) -q/\302\245
= 2c'tan(45
\302\260
+ \177)'/2)/y-q/y (8.12)

The total active thrust may be corrected by ignoring the stress distribution in the tensile zone.

Example 8.1
Calculatethe total active thrust per metre length of an 8 m high smooth vertical retaining walI,
The properties of the backfill soil are: c' = 20 kPa, \177b'= 25 \302\260,
and y = 17.5 kN/m 3,

Solution:
\302\260 = tan:Z(45.0 \302\260
ka = tan\177-(45.0 -#'/2) -25.0\302\260/2) = 0.406.)))
 Lateral Earth Pressure and Retaining Walls 309

XHk a
- 2c\" k.'.\177a qka (yH + q)ka
- 2c\177

(a) (b) (c)

Fibre 8.4. Distribution of active lateral e\177h pressure behind a smith retaining wall.

At z = 0, o\"z = 0 and Pa = -2c\" = -40,0 0.\177\"\177 = -25.5

\177'\177-\177'a
kPa.
At z = 8,0, \177z = 8,0 x 17.5 = 140.0kPa, Pa = 140.0x0,406- 40.0 0,-f\177-\177 = 31,3 kPa.
The depth at which the active pressure becomeszero is calculated using Equation 8.12:
x0-\177.406)
z o = 2 x 20.0/(17,5 = 3.59 m. Ignoring the tensile force:

Pa
= (8.0- 3.59)x31.3x 1.0/2= 69.0kN. The point of application of this force is:

(8.0 - 3.59)/ 3 = 1.47m above the base of the wall (Figure 8.5).

Example 8.2
A smooth vertical retaining wall is 5 m high. Determine the total active thrust for the follow-

ing cases: (a)_there is no hydrostatic water pressure in the bacld:ill: c' =0 \177\"= 36 \302\260,
and
=
\302\24517.5 kN/m 3, (b) the water table is at the ground surface: c = 0, O = 36 , \177tsat= 19.5 kN/m .

Solution:
For both cases: k = (1-sin36.0\302\260)/(l+
sin36.0 \302\260)
=0,260,
a

-25.5 kPa

4.41/3 = 1.47m)))

31.3 kPa

Figure 8.5. Example 8.1.

310 Soil Mechanics: Basic Conceptsand Engineering Applications

(a) At z = 0, \177\":= 0, and Pa = O,

At z = 5.0 m, o\"z= 5.0 x 17.5 = 87.5 kPa, and Pa
= 87.5 0.260
\303\227 = 22.7 kPa,
Pa
= 22.7xS.0/2x 1.0 = 56.8 kN.

(b) At z = 0, o\"z= 0, and Pa = 0. At z = 5.0 m, pore pressureu = 5.0\303\2279.81 = 49.0 kPa,

cY\177
=5.0x19.5-49.0 = 48.5kPa, Pa = 48.5x0.260= 12.6kPa.
= [2.6\303\2275.0/2 x 1.0 = 31.5 kN.
P,

P\177.
= force due to water pressure
= 49.0 x 5.0/2 x 1.0= 122.5
kN,

Total thrust on the back of the wall = Pa+ P,,, = 3 1.5+ 122.5= 154.0
kN.

earth pressure = O, d?\" soil

8.3.3 Linear distribution for a sloping ground surface in c\"

Figure 8.6(a) illustrates a retaini\177lg wall with the sloping grouM surface at an angle 13 to the
horizontal. The soil is granular with shear strength parameters c' = 0, 9' and 13 < \177'. To evalu-
ate the active earth pressure pa at depth z consider a rhombic element of soil of width of dt. On
the vertical sides of the element the stress is pa and is assumed to be parallel to the ground sur-
face. This implies that the wall has a friction angle of 13 with the backfill soil. The vertical

stress at depth z is equal to the weight of the material in the volume abed divided by ab:

or'. =(dl x z x 1.0x y)/(dI / cos [3) = yz cos 13

Thus the normal and shear stresses on plane ab are:

= 2 =
cry, b yzcos [3, \"tab yzcos[3sin [3 (8.13)

The states of stress on plane ab and the vertical plane are representedby points B and A re-
spectively on the active stress circle shown in Figure 8.6(b). The length OB is equal to the ver-

\177
T1\177n6cos

=
\177'z yzcosl\177

Figure 8.6. Active state behind a wall with sloping backfill in c\" = 0. 0\" soil,)))
 Lateral Earth Pressure and Retaining Walls 311

tical stress on plane ab: similarly OA represents Pa. Numerically OA\" = OA, and therefore from
the geometry of Figure 8.6(b), tl\177e earth pressure coefficient is given by:
z
_ Pa _ OA' = OC- A'C - cos\177- \177/cos \" f\177- cos 0'
ka
k
\303\267
a (8.14)
OB OC + CB 2
\177
cos13+ \177Jcos 213 - cos \177'
The active earth pressure at depth z is:

cos[3 (8.15)

and the total active thrust that acts at height H / 3 above the base is given by:

G = 0.5\177,hr
2
G cos 13 (8.16)

Using a similar procedure the following relationships for the passive state can be obtained:
z
cos13 + \177Jcos: 13-cos O'
\177P
(8.17)
cos13
- e
Ig -co.s
\177
0\"
\177/cos

The passive earth pressure at

depth z and the total passive thrust are calculated using equations
silnilar to Equations 8.15 and 8,16 in which ka is replaced by kp. In the presence of a surface
gravity load of q, yz must be replaced by yz + q,

Example 8.3
A wall with a vertical back is 5 m high and retains a sloping soil with
retaining 13
= 20 \302\260.
De-

termine the magnitude of the active thrust and the inclination of the failure planes developed
behind the wall. c' = 0.4}' = 36 \302\260,
and y = 17.5 kN/m 3.

Solution:

Using Equation 8.14 gives ka = 0.326. The total thrust from Equation 8.16 is:
-\177
0\"5x17'5x5'02 = 67.0
x0\"326xc\302\260s20'0\302\260xl'0 kN.
Pa

Calculate the state of stress on a rhombic element at an arbitrary depth (say 5 m):
\177r'
z
=
yzcosl3
\302\260
= 17.5\303\2275.0xcos20.0= 82.2 kPa, Pa = YZka cos[3 =82.2\303\2270.326= 26.8 kPa.
From the circle of Figure
stress 8.7: A'C = A'B/2 = (82.2-26.8)/2 = kPa, 27.7
\302\260
OC = 26.8 \177- 27.7 = 54.5 kPa, OC 1 = OC/cos[\177 = 54.5/cos20.0 = 58.0
kPa,
\302\260
= x sin qb' -- 58.0 x sin 36.0 = 34.1 kPa.
C\177T\177 OC\177

) = CB/ CI B = A\" C / C\177T\177= 27.7 / 34.1 --\177ZCBC\177 = 35.7 ,

\302\260
cos(ZCBC\177 \302\260 \302\260
B = + 13 = 35.7 + 20.0 = 55.7 \302\260.
ZSC\177 ZCBC\177
The angles between the inclined planes of the rhombic element (representedby point B), with
the failure plane of T1 and T2, are 0:t and 0:2 respectively:
= - = \302\260 =
o\177 1/2(\177SC\177T\177 \177SC\177B) 1/2(90.0\302\260+ 36.0 -55.7 \302\260)35.2 \302\260.

\302\260 \302\260 \302\260

Thus of failure
the angle plane TI from the horizontal 35.2
is: + 20 = 55.2 (anticlockwise).
= 1/2(\177SCT - ZSC B) = 1/2(180.0 \302\260
+ 90.0
\302\260
- 36.0 \302\260
- 55.7 =
\302\260)89.2
\302\260
.
cQ I 2 l
\302\260 \302\260 \302\260
Consequently, the angle of failure plane T2 from the horizontal is: 89.2 + 20 = 109.2 (anti-
clockwise). The results are presentedin Figure 8.7. It may be shown that the failure planes T I)))
3 i2 Soil Mechanics: Basic Conceptsand Engineering Applications

77.3 kPa
[\177

Figure 8.7. Example 8.3.

and T2 are parallel to the lines ATI and AT2 respectively. Note that these directions are
independent of the unit weight and depend entirely on the \177' and 13 values.

8.3.4 Stratified backfill

At the boundary of the two soils in a stratified backfill, two lateral earth pressures exist, each
calculatedusing the corresponding earth pressure coefficient. The total vertical stress is always
linear with depth but its slope changes
change with
weight. the in vertical
unit The effective
stress has a similar pattern water table is if the
continuous within the soil. If water is resting on
an impervious layer located behind the wall, then for the pervious layer the effective vertical

stress is used whilst for the impervious layer the total vertical stress is used.

Example 8.4
A retainiag wall 6 m high supports two layers of soil each having a thick,hess of 3 m. The
\177
c = 0, e =
\" \" \302\260
properties of the layers are: 30 , \"\177d\177
17.5 kN/m , and = 19.5
\177sa\177
upper,layer: 3--
kN/m3; lower layer: c' = 10 kPa, 0 and Y\177at = 19 kN/m . There is a surfaceload
= 18\302\260, of 50
kPa and the water table is 1.5 m below the ground surface. Determine the total lateral thrust
and its line of action above the base of the wall.
Solution:

For the upper layer: ka = (1 - sit\17730.0 \302\260)/(1 = 0.333,

+ sin 30,0 \302\260)

For the lower layer: k = (1 - sin 18.0\302\260)/(1 = 0.528.)))

+ sin 18.0 \302\260)
a
 Lateral Earth Pressure and Retafning Walls 313

At z = 0, \177z= 50.0 kPa, u = 0,

o-':= 50.0 kPa, Pa = (Y'zka = 50.0 \303\227
0.333 = 16.7 kPa.
At z = 1.5 m, \177z= 50.0 + 1.5 x 17.5 = 76.2kPa, u = 0, o\":= 76.2 kPa,
Pa = 76.2x 0.333= 25.4kPa.
At z = 3.0 m, \177r z 19.5 = 105.5 k.Pa, u = 1.5 x 9.81 = 14.7kPa,
= 50.0 + 1.5 x 17.5+ 1.5\303\227
= - 14.7 = 90.8 kPa, ka = 0.333, Pa = 0\177:ka = 90.8 x 0.333=30.3 kPa.
c\302\242.:105.5 using

Usingka -- 0.528,
Pa
= -2c'x =90\"Sx0'528-2xi0'0\177 = 33.4 kPa.
cY'zka k\177a

At z = 6,0 m. cE= 50.0+ 1.5x i7.5+ 1.5 x 19.5 + 3.0 x 19.0= 162.5kPa,
=
t\177 4.5 x 9.81 =44.1 k_Pa, o\"z
= 162.5 -44.1=118.4 kPa,
Pa = 118.4x0.528
-2x10.0\177 = 48.0 kPa.

Note that due distribution of the lateral pressure with depth we can find the lateral
to the linear
earth pressure at the bottom of any saturated layer by adding y' x h x ka (oryx h \303\227
ka with no
water in the system) to the lateral pressure at the top of the layer that has already been calcu-
lated. The term y' represents the submerged or buoyant unit weight. The results are shown in

Figure 8.8 and the computations are summarized in the table below.
Distance of the total horizontal thrust from the base = 640.85 / 294.66 = 2.17m.

Arm above Force \303\227arm

Force/kN)
the base (m) (kN.m)
FI
= 16.7x 1.5x 1.0 (per meter run) = 25.05 5.25 131.51
F2= (25.4- 16.7)x 1.5x i / 2 x 1.0 = 6.52 5.00 32.60
F3= 25.4 x 1.5 x 1.0 = 38.10 3.75 142.87
Fz = (30.3
- 25.4)x i.5x I / 2 x i.0 = 3.67 3.5 12.84
F5
= 33.4 x 3.0 x 1.0= 100.20 1.5 150.30

F6= (48.0- 33.4) 3 x

\303\227 1/2x 1.0=21.90 1.00 21.90
F7 = 44.1 x4.5 1/2
\303\227 x 1.0 =99.22 t.5 148.83
Total Thrust: 294.66 Total: 640.85

16.7

F2
WT
.4
\177 F4 .5 m
\177,\177,\177

30.3 33.4
I,\177

F5
_.\177 3.0 rn
i\177F6

48.0 44.1
All stresses in (kPa)

Figure 8.8. Example 8.4.)))

314 Soil Mechanics: Basic Concepts and Engineering Applications

8,4 COULOMB WEDGE ANALYSIS

8.4.1 Principlesof the Coulomb wedge analysis

The Coulomb wedgeanalysis is a limit equilibrium method. A wedge shape collapse (or fail-
ure) mechanismis assumedwhich is bounded by the back faceof the retaining wall, a horizon-
tal or inclined, loaded or unloaded, ground surface and a linear failure plane (Figure 8,9(a)).
The wedge mechanism slides downwards or upwards on the failure plane simulating the active
and passive states respectively. The aim of the analysis, which is based on the force equilib-
rium of the wedge, is to obtain the critical value of the sliding angle \177 and the corresponding
active or passive thrust. Along the sliding surface, the Moltr-Coulomb failure criterion is as-
sumed to apply. For a granutar material the wedge and corresponding force diagram for the ac-
tive state are shown in Figures 8.9(b) and 8.9(c) respectively. In the case of a granular material

with no surface loading, it is possible to formulate the problem (obtaining the values of active
and passive thrusts) mathematically without a need for iteration, However, with the inclusion
of cohesionon the sliding surface and on the back face of the wall, an irregular ground surface
or surface loading, an iterative procedure must be used, Note that the soil reaction R and the
active thrust Pa are both located under the line perpendicular to the direction of sliding. This
ensures that the senses of the components of these forces along the sliding surfaces (shearre-
sistance) are opposite to the movement of the wedge. If the base of the retaining wall is con-
structed on a soft soil and the settlement is predicted to be higher than the displacement
neededfor active failure, then the force Pa may be relocated by making an angle --6\" with the
line perpendicular to the back face of the wall. A graphical representation of the iteration of
the angle c\177,and locating its critical value and the corresponding maximum active thrust, is il-

lustrated in Figure 8.10. The weight of each trial wedge W is computed from the geometry of
the wedge and all such weights are plotted on a common vertical load line (a process known as
Culmann procedure). For each trial wedge in which c\177 is known, the angle between W and Pa
is 180 - 0 - \1775'and the angle between W and R is ct - \302\242p'.
\302\260
It i'ollows that two lines representing
forces Pa and R can be drawn from both ends of the load line so that they intersect thereby
giving the magnitude of Pa for each trial.

180\302\260-;0 + # C_........ C

-e-\177'
]\" \177 -\", ,' \177--I\177 \177 I /

R
B B

Ca) (b)

Fibre 8.9. Coulomb w\177ge analysis in c' = O. @' soil (active state).)))
 Lateral Earth Pressure and Retaining Walls 315

(a) (b)
Figure 8.10. Graphical presentation of wedge analysis in c' = 0, \177\"soil (active state),

The maximum value of Pa and the corresponding triangle of forces are found by drawing a
vertical tangent to the curve passing from the intersection points described above. If there is a
uniform vertical surface loading (q) on the ground surface then the weight of each wedge is
increasedby q \303\227
AC \303\227
cosl3. In the case of a line surcharge load it is convenient to locate a
trial failure plane passing through the point of applicatiot] of the load as shown in Figure
8.1 l(a). The corresponding wedgeis considered twice, once without the surface load and once
with the surface load and there will be a jump on the Culmann li\177]e at this point. However the

maximum Pa is found in the same way as explained previously. The surface load is included

W3 + Q\177

paa.\177,\177\"J
/

C2 C C3 W2+Q1
\177
C1\177
A

W\177
+
Q\177

W\177

\177\"
. L2_ ............ _,e
B

(a) (b)

Fibre 8,1 \177.Graphical presentalion of wedge analysis: effect of a line surch\177ge load in the active state.)))
316 Soit Mechanics: Basic Concepts and Engineering Apptications

in the weights of the wedges thereafter. The above procedure is easilyadapted to include the
cohesion resistance developedon the sliding surface and the back face of the wall. For the pas-
sive failure where the wedge is pushed upwards, the inclination of forces R and Pp must be se-
lected so as to hold down the wedge and create the shear resistances in the opposite direction
of the movement of the wedge. Note that in the passive state we are searching for the smallest
force that can move the wedge upwards.
When there is water behind the retaining wall, forces due to the water pressure on the
boundaries of the trial wedge (BA and BCI, BC2....) must be included in the force diagram.
The force diagram will yield the effective active or passive thrusts. It can be shown that if the
total weight of the wedge is considered along with water pressure on the BC\177, BCz,...edges

and the water force on the BA edge is deleted, the force diagram will yield the total thrust.
The effect of wall friction near the base of the wall is to produce a curved failure surface

(Figure 8.12). Assuming a linear failure surface introduces errors in the estimation of the ac-
tive and passive thrusts. In the active state the error is negligible, but the passive thrust will be

significantly overestimated depending on the value of the mobilized friction angle on the back
face of the wall 03') and the wall displacement. In the traditional method of evaluating the pas-
sive thrust it is assumed that the failure surface immediately behind the base of the wall is ei-
ther a logarithmic spiral or a circle. As the distance of the failure surface from the wall is
\177
creased it becomes linear and makes an angle of 45 - 0' / 2 with the horizontal. The analysis
is carried out in two steps and begins by ignoring the cohesion from which a minimum value
of the passive thrust is obtained. The point of application of this thrust is assumed to be H / 3
above the base of the wall. In the second step the passive thrust due to cohesion and friction is
calculated by neglecting the weight of the material. The point of application of this thrust is

assumed to be H / 2 abovethe base. The failure surfaces from these two steps may not be iden-
tical; the results are combined to give the total passive thrust located between H / 2 and H / 3.
A difference of approximately 10% is expected with the results obtained using linear failure

plane and a triangular wedge (Terzaghi et al., 1996).

Active Passive
failure failure

\177
Assumed
\177z failure plane
Pp

-
// Assumed

Pa
plane
\177\177-\177
,f\177failure
Probable failure Probable failure
surface surface
(a) (b)

Figure 8.12. Development of the active and passive failure surfaces due to wall friclion.)))
 Lateral Earth Pressure and Retaining Walls 317

The magnitude of 8' dependson the angle of internal friction of the backfill soil and the
roughness of surface. For a concretewalt
the wall a value between /
\177b' 2 and 2\177' / 3 is used ex-
tensively in design. Values can also be determined in the laboratory using direct shear type
tests on composite concrete-soil samples. In this way an estimate can be made of the adhesion
developed between the wall material and the backfilI soil.

8.4.2 Coulomb wedge analysis for c\" = O,

\177' soil

Referring to Figure 8.9(c), equilibrium of the forces results in:

+ - - sin(o;-
sin[(0 \1773') (o\177 \177?')] 0')
where W = SAeC X 1.0 X \177is the weight of the trial wedge, and Pa is the active thrust corre-
sponding to the trial angle \177. The \177ea of the wedge, calculated from its geome\177, is given by:

J J

Substituting W into the force equilib\177um \177uation we obtains:

A maximum value for ka can be found by setting aP a/a\177

= 0, Thus, for the active case:

sin 2 (8- \177')

ka= (8.19)
+ \177)sin(\177,_ \177)
sin2Osin(O+8 ') /sin(,,
[ i+\177 ]2

Far a ve\177ical wall (a = 90*) and a horizontal gound surface (\177

=
0):

cos 0'
\177\"
= (8.20)
\177 + 4sin(0' + 6')sin,'

For a smooth (\177'

= O) veaical wall with horizontal ground su\177ace, ka becomes identical to
Equation 8.6. Coulomb's theo\177 is also used to establish the passive soil pressure coefficient
using the linear failure surface. By considering forceequilibrium of the passive wedge shown
in Figure 8.13(a), and using the force diagram of Figure 8.13(b), tim following expression is
dete\177ined:
2
sin (0 + \177')

\177
sin \177sin(\177-\177') 1
\177)))
318 Soil Mechanics: Basic Concepts and Engineering Applications

_C

\177-(\177+\177)

(a) (b)

Figure 8. ] 3. Coulomb wedge \177a]ysis in c\" = O, \177\"

soil (passive state).

For a vertical wail (13

= 90 \302\260)
and a horizontal ground surface (13 =
0), Equation 8.21 becomes:

cos\177'
= (8.22)
kp _
'.\177'\177--\1777 ,\177sin(\177)\"+ \177')sin \177\"

For a smooth (6' = 0) vertical wall with a horizontal ground surface, this equation reduces to

Equation 8.9.

8.4.3 Coulomb wedge analysis in c,, TM

\177u 0 soil

Undrained conditions in both the active and passive states may occur behind a retaining wall
and the analysis is representative of a short-term stability. The forces acting on the wedge are
shown in Figure 8.14(a), where Tc and Tw are the shear forces developed on the sliding surface
and wall-soil interface respectively.In the active state, tension cracks are permitted to extend
to a depth zo that can be evaluated in terms of the undrained cohesion c, and wall-soil adhe-
sion cw. The/atcral thrust due to the water pressure in the te\177sion cracks (Pw) may be taken

into account separately. In the development of an analytical solution we first consider the ac-
tive case for a vertical wall with cohesion, and a horizontal ground surface. The sliding surface
terminates at the bottom of a tension crack and the reaction force R and shear resistance Tc are
both mobilized along the length BD. Figure 8.14(b) shows the force diagram excluding P\177..
Equilibrium of the vertical and horizontal components of the forces yields:
T
c sita\177 + T w + Rcosc\177-W=0
Tc cosc\177- Rsina+ Pa
=0

Combining theseequilibrium equations and replacing Tw, Tc and W with T\177v

= c (H
w -z o),
T
c
= -
c, (H zo)/si\177a\177,W
= \177-
\302\245(H zS,)cot c\177/2, and setting cqPa/c3c\177 =0 we obtain:

coto\177 = + c w/c u (8.23))))

\1771
 Lateral Earth Pressure and Retaining Walls 319

Pa

A E C

'W|

B
'T

(b)
Figure Cou[omb's
8\303\26714. wedge analysis in c. %= 0 soil (active state).

2_ -
Pa
=
y(H zg) / 2 -
2c. (H z o )\177l + c\177 I c,for q =0 (8.24)
If vertical surface loading q exists then:
Pa =\177'( He
-zZo)/2+q(H-zo)-2c,,(H-zo)\177l+c.,/c,
for q >0 (8.25)
Note that the magnitude of zo is unknown but may be estimated from Equation 8.12. Equation
8.24 is equivalent to a linear distribution of lateral earth pressure given by:

Pa = yz
- 2c + cw / cu
u \177]1 (8.26)

where z = 0 at the ground surface and z = H at the base of the wall. With the assumption of a
linear lateral pressure distribution, the depth at which the earth pressure becomes zero is:
=
z o (2c. / \302\245).\1771
+ c w/c u (8.27)

The total active thrust is the integral of Equation 8.26 with a correction for the tension zone:

ea
= \177H (H
- z / 2-
o) c,,(H
- z o )\1771+ c w I c,(8.28)
q > 0. and Pa < 0 at z
For = 0, the linear lateral earth pressure, of tension
equivalent depth
crack and total active thrust are given by:

p, = yz + q -
2% \177]1+ c\177./ c, (8.29)

zo = (2c. + cwI c u - q/y (8.30)

/y)\177l

Pa =yH(H-Zo)/2+q(H-zo)-cu(H-zo)\177/t+cw/c u (8.31)

The foregoing analysis for the active state can be extended to include an inclined wall and

sloping ground surface. However, due to the complexity of the mathematics, an iterative)))
320 Soil Mect\177anics: Basic Concepts and Engineering Applications

method using the variable 0t is preferred. For the passive state, the direction of the forces Tc
and T\177.are reversed and, using a similar analysis, we obtain:

Pp =yH2
/2 +qH + 2cuH41+cw/c
u (8.32)

The critical magnitude of \177is given by Equation 8.23.

Example 8.5
An 8 _m high wall retains a soil with
.
the following properties: c, = 35 kPa, \177)u
= 0,
),
= 19.5
3
kN/m , and c\177v= 16 k_Pa. Determine the magnitude of the active thrust Pa \342\200\242

Solution:

From Equation 8.27: z o = (2 \303\227

35.0 / 19.5)a/1 + 16.0/35.0 = 4.33m.
Using Equation 8.28) to compute
8.24 (or
Pa
= 19\"5(802 -4'332)/2-2\303\22735\"0(8\"0-433) x/1+16\"0/350=131.1kN.
From Equation 8.23 the corresponding wedgeangle ct at failure is:

coto\177 = 41 + 16.0/35,0 = \177.207 -\177 c\177= 39.6 \302\260

. The force due to the hydrostatic water pressure
the tension crack is: P,,. = (1 / 2)9.81\303\2274.332= 92.0 kN.
Hence, the total horizontal thrust = Pa + Pw = 131.1 + 92.0 = 223.1kN.

8.4.4 Coulomb wedge analysis in c', soil

\177\"

The trial, wedge for the active state, together with the corresponding forces and the force
diagram, is shown in Figure 8.15. Tension cracks are allowed to extend to the depth Zo that
may be estimated from Equations 8.12 or 8.30. As before, these cracks may be filted with
water (from rain for example) and so the force due to the hydrostatic water pressure Pw must
be taken into account. For the passive failure condition where the wedge is pushed upwards,
the inclination of the forces R and Pp must be selected so as to hold down the wedge and
create the shear resistances in the direction opposite to the movement of the wedge. Similarly,
the direction of the forces Tcand Tw are reversed.

Example 8.6
A vertical retaining wall has a height of 8 m with ground surface sloping upward at an angle of
15 \302\260.
Determine the active thrust Pa due to trial wedges of soil having planes at 45\302\260and
!,ailure
60
\302\260
to the horizontal, c' = 10k_Pa, O\" = 25\302\260,
Y
= 18 kN/m 3, cve = 0, and ,5 = 20\302\260.
Solution:

The forces acting on the trial wedge ABDE are shown in Figure 8.16(a).
\302\260
From Equation 8.12: Zo
= 2\303\22710.0\303\227
tan(45.0 + 25.0\302\260/2)/18.0= 1.74 m.
Calculatethe area of block ABC from the geometry of triangle ABC:
\302\260
AB / sin((z-15.0 \302\260)
= AC
/sin(90.O -(z) --\177 AC = ABcosct/sin((z-15.0\302\260).
=
GC ACcosl5.0\302\260, GC = AB cos ct cos 15.0\302\260/sin(or - \1775.0\302\260).
A reaA\1773C = GC / 2
AB \303\227 = AB 2 cos ct cos 15.0 \302\260 / 2 sin(o: - 15.0\302\260),similarly,
2
= ED cos\177cos 15.0 / 2 sin(or - 15.0\302\260).Calculate
\302\260
the weight of the W:)))
AreaED C wedge
 Lateral Earth Pressure and Retaining Walls 321

-
\177
0 - 6'

E \177
P\17780

(a)

(b)

Figure 8.15. Coulomb' s wedge analy sis in c', 0' soil (active state).

W = - = 530.0533cos -
(AreaA\177 C AreaEDC) Y, W \177/sin(c\177 15.0\302\260),

For W =
c\177= 45 \302\260. 749.6 kN, and for o. = 60 \302\260,
W = 374.8 kN.

The lengths BC and DC are calculated from triangles ABC and EDC:
BC = ABsin 105.0\302\260/sin((x
-
15.0\302\260),DC
= EDsin 105.0'\177/sin0x- 15.0\302\260), thus:

a
Pa
h
\177, 90 -
o ,5\"
I

90 \302\260
+ I

w\177

(a) (b)

Figure 8A6. Example 8.6.)))

322 Soil Mechanics: Basic Concepts and Engineering Applications

-
/ 1Z_&___. .\177IZ_'\177_*\177__.

Pa B B Pa B
(a) (b) (c)
Figure 8.17. Location of the point of application of the active thrust.

Tc = (BC- Tc =
DC)\303\227l.O\303\227c\"--\302\242, 60.467 / sin(\177 - 15.0\302\260).

For c\177= 45 \302\260, Tc =

To= 120.9 kN, and for ct = 60 \302\260, 85,5kN.
Calculate: = 1.74xl.74\303\2279.81/2 =14.8 kN, the force of Figure 8.16(b):
Pw Using diagram
For ct = 45\302\260:

dg
= Tecosct - 14.8= 120.9\303\227
cos45.0 - 14.8 = 70.7 kN.
\302\260

ce = dg / cosct = 70.71cos45.0= i00.0 kN.

\302\260

In triangle bec:
i00,0 / sin(o\177 -
\177')
= bc / sin(90.0 +
'\177
\177') --) bc = 265,0 kN.
ab = W- bc - cd = 749.6 - 265.0- i4.Stan45.0
\302\260
= 469.8 kN.
- \302\260
in triangle abh: Pa/ sin(cx \177')
= ab/sin(90.O + 5'- + \177'),
o\177

\302\260
- \177' \302\260
- \302\260
25.0 = / sin(90.0 45.0
Pa/ sin(45.0 \302\260)469.8 + 20.0 + 25.0\302\260),

= i60.7 kN.
Pa
For \177,= 60 ce
\302\260,
= 55.8 kN, bc = 88,2kN, ab = 261.0 kN and Pa
= 155.0 kN.

8.4.5 The point of application of the active thrust

The precise determination of the point of appIication
active of
thrust remains a difficult the

task despite significant improvements and experimental techniques in the

in the theoretical

analysis of retaining structures. This is mostly due to uncertainties in the lateral earth pressure
distribution along the wail. The traditional method used to locate the point of application of
the active thrust is shown in Figure 8.17. In the absence of any surface line load, the point of

application above the base (point D in Figure 8.17(a)) is 1/3 of the height of the wall. If,
however, a line load is located within the critical wedge, then the point of application moves
upwards so that the length FD = 1! 3 FK. The lines EF and EK are parallel to the lines BB' and
BC respectively where BB' makes an angle of \177' with the horizontal (Figure 8.17(b)). If the
li\177e load is located outside the critical wedge (Figure 8.17(c)), the length FD = 1/3 FB where
EF is parallel to BB'.)))
 Lateral Earth Pressure and Retaining Waits 323

8.5 COMMON TYPES OF RETAINING STRUCTURES AND FACTOR OF SAFETY

8.5.!. Cot\177ventional or externally stabilized systems

Retaining structures may be classified of ways. O'Rourke & Jones (1990)pro-
in a variety

posed a classification retaining of two

structures into
major groups, vizL, externally and inter-
nally stabilized systems. An externally stabilized system uses a structural watt external to the

soil as in the conventional retaining walls shown in Figure 8.18. An internally stabilized sys-
tem, on the other hand, involves reinforcement installed within the backfill soil.
The gravity type wall of Figure 8.i8(a) may be constructed fro\177 brick, masonry and plain
or reinforced concrete.Both the front and back faces may be vertical, inclined or stepped.
When the base and stem are designedand constructed as separate elements, a shear key must
be included in the stem to increase the factor of safety against sliding. Similar shear keys are

designed to increase the frictional and adhesion forces acting on the interface between the base
and the foundation soil. The stability of a gravity wall is maintained mostly by its weight and

partly by the passive resistance mobilized by the soil at the front of the wall. Assuming the
wall remains intact, failure occurs by horizontal displacement or by rotation about the toe.
Therefore factors of safety must be determined for both sliding and overturning of the wall.
Excessive contact pressure and settlement of the toe may cause a shear failure of the founda-
tion soil. Consequently, it is necessary to evaluate the \177tability of the wall and soil mass as a
whole using the slope stability methods. Cantilever walls of the type illustrated in Figure

8.18(b) are typically coustructed from reinforced concrete. Such a wall is either L-shaped or
T-shaped and is considered to be a flexiblestructure. L-shaped prestressed concrete cantilever
walls are also utilized.The projection of the base (the heel portion of the wall) must be posi-
tioned inside the backfill. Shear key(s) are constructed on the cut off points and the base of the
wall. Stability of the wall is maintained by its own weight and the weight of the retained soil
above the heel. Dependingon the height of the soil, any mobilized passive resistance at the

front of the wall may also be takeu into account. Long cantilever retaining walls are often sup-
ported along their \177ength by counterforts to increase their rigidity and strength. The thickness
of the base is normally governed by an allowable shear stress in the concrete; reinforcement is
provided in zones of tensile stress aud to control shrinkage cracks. Figure 8.I8(c) represents a

gravity type retaining wall called gabion walt constructed from cubic (or other shapes) baskets
made of metal wire or hard plastic (geogrids) and filled with coarse aggregate. If the soil be-
hind the wall is also reinforced by geogrids, structures as high as 8 m may be built using this
technique. A gabion wall may be vertical or inclined dependi\177lg on the angle of internal fric-

tion of the aggregate. If the grading of the aggregate complies with the specification of a filter
material then there is no need for the construction of a drainage system. Excessive contact

pressure under the base of a gabion wall generates settlement that may lead to a gradual shear
distortio\177 a\177d lateral deformation of the wall (O'Rourke. 1987). Coutact pressures may be re-
duced by reinforced solutions. Applications of gabion walls are spreading with the prolifera-

tion of dual*purposeproducts (reinforcing and filtering) on the market (Pfilossy et aI., 1993).A
general layout of a crib wall is shown in Figure 8.18(d). It is composed of precast concrete
elements that are either solid or hollow. Hollow elementsmay be filled with soil to provide)))
324 Soil Mechanics: Basic Conceptsand Engineering Applications

h--- h--
(c)
(a) (b)

cam Anchor rod m

H
ie
WT
Anchor

(d) (e)
Figure 8.18. Conventional types of retaining structures: (a) gravity wall, Co) cantilever wall. (c) gabion
wall, (d) crib wall, (e) cantileyer or anchored sheet pile.

weight and improve stability. Construction of this type of retaining structure is rapid, as it does
not require special skills. Similar to a gabion wall, there is no need to provide a drainage sys-

tem and therefore this form of construction has an economic advantage over other types of re-
taining structures. Recently, the longitudinal beam elements at the rear of crib walls have been

ips or grids

t
Facing/
\"
panels
H ,, H

(a) (b)

Figure 8. [9. In\177erna][y stabi]ized ean\177 retaining structures: (a) reinforced soil (b) soil nailing.)))
 Lateral Earth Pressure and Retaining Walls 325

omitted to improve economy.If the crib wall is supported by an anchorage system its effec-
tiveness becomescomparable with gravity and cantilever walls.
Sheetpile systems (Figure 8.18(e)) are mostly temporary retaining structures that are built

to facilitate excavations or retain soils. However, the use of permanent sheet pile systems is
common in both onshore and offshore pile is a flexiblebeam
structures. A sheet
constructed
from concrete, timber or, most commonly, steel. Whilst precast concrete piles are quite heavy
and difficult to handle and drive, they may be competitive with steel piles if they are cast close
t6 the jobsite. Timber sheet piles are for low heights up to 3 m. The stability of a sheet pile
wall is maintained by embedment depth using the mechanics of a cantilever beam. When the
embedment depth is not adequate or decreases during excavation, anchors may be used to in-
crease the stability.

8.5.2 Improved earth walls or internally stabilized systems

Improved earth walls consist of stabilized backfill soil and facing elements. Improvement of
the soil (apartfrom compaction) is carried out either by means of chemicals or by using inclu-
sions. Chemical stabilization of the soil may be realized by cement or lime or other chemicals,
Soils stabilized by cement or lime normally fall into the category of a Mohr-Coulomb material
with improved shear strength parameters. A mechanically stabilized soil is reinforced by strips
or grids (Figure 8,19(a)) that may be metallic, polymeric or organic. A mixture of soil and
polymeric elementsof fine diameter and small length has also been used. The main objective
is to transfer the tensile stresses to reinforcement elements. Anchored earth systems combined
by soil reiuforcement have been developed and applied successfully in highway construction.
The in-situ reinforcement includes soil nailing and dwelling by means of grouted bars as
shown in Figure 8.19(b).
In the selection of a retaining wall for a specific project, considerationshould be given to
the type of soil, its deformation compatibility with the retaining wall, the height of the retained
soil, ground water level, construction and environmental aspects, time and cost.

8,5.3 Factor of safety in retainfng walls

For externally stabilized systems the factors of safety against sliding and overturning are con-
sideredseparately. The factor of safety against sliding is defined as the ratio of the sum of the
resisting forces to the sum of the disturbing forces along the sliding surface:

FS
- \177F\177.
(8.33)
EF d
Forcesthat resist sliding include of at the front
the of the wall, and adhesive and
passive thrust
frictional forces mobilized on the sliding The disturbing forces are the components
surface. of
the active thrust in the direction
slidingof the
and the force due to any water pressure behind
the wall. For overturning, the factor of safety is defined as the ratio of the sum of resisting

moments to the sum of disturbing moments about the toe of the retaining wall.)))
326 Soil Mechanics: Basic Concepts and Engineering Applications

F v
r (8.34)
\177M d
In most cases the adhesive and mobilized under the base have no effect as
frictional forces
they pass through the toe. The contact pressure under
resultant of the the base N, which is

equal to the total vertical forcesapplied to the wall, has no moment about toe because at this
limiting state it approaches the toe as the wall loses contact with the foundatioll soil along the
base. For both sliding and overturning, the factor of safety must be not less than 1.5; reference
should be made to the standard codes of the relevant country.
For a reinforced earth wall a simplified traditional analysis assumes that the active thrust in
the soil is fully mobilized through a linear distribution along the back face of the wall. This
lateral active pressure is resistedsolely by the strips or geogrids so (theoretically) there is no
pressure on the facing elements. The factor of safety is applied to the tensile strength of the

strips and geogrids takes due account of environmental and construction factors.
The factor of safety in a sheet pile system is appliedto c', \177', the passive forces or kp. For
strength data of moderate retiability, a factor of safety of 1.3 to 1.5 is applied to both the cohe-

sion and internal friction angle but for less reliable data a factor of safety of 2 is normally
adopted.

8.6 STATIC ANALYSIS OF CANTILEVER AND GRAVITY RETAINING WALLS

8.6.i Soilpressuredistribution under the footing

The contact pressure under the footing is assumed to be linear and, ideally, compressive
throughout. Maximum contactpressure occurs under the toe, while the minimum pressure oc-
curs under the heel. If the contact pressure becomesnegative at the heel then the factor of
safety against overturning must be calculated ignoring unreliable properties such as wall fric-
tion and adhesion. In this case a part of the contact area becomes ineffective and the computed
toe pressure increases and may exceed the allowable bearing pressure.

8.6.2 Basic concepts of static analysis

Typical sectionsof gravity walls are shown
cantilever and in Figure 8.20, together with the
forces applied to the weight of the watt W is comprised from the materials
walls. The within
the dotted lines. Thus the active thrust is calculated on the imaginary wall ab assuming that the
soil at the left of ab is a rigid material attached to the rigid wall. Similarly the passive thrust is
calculated on dc. The shear force Trepresents the adhesion and frictional forces mobilized un-
der the base whilst the normal force N is equal to the sum of the vertical forces applied to the
wall system:

T = l--\177-(Bc b +Ntan\177b), N =W +Par \) +Ppv+Q (8.35)

Fs
 Lateral Earth Pressure and Retaining Walls 327

(a) (b)

Figure 8,20. Sections and flee-body diagrams of typical cantilever and gravity retaining walls.

where cl\177and *hb are the maximum values of adhesionand friction mobilized under the
angle
base respectively, Fs is the factor of safety against sliding, Pay is the vertical component of the
active tbxust, Pp,, is the vertical component of the passive thrust and Q is the resultant of the
vertical surface load appliedat the upper boundary of the dotted lines. It is assumed that all the
resisting forces have been decreasedby the ratio 1 / Fs to maintain equilibrium,
To calculate the active thrust on the imaginary wall ab, the concept of a linear earth pres-
sure distribution can be employed in/he following form:

= (7z + q)k - + c w/c')

Pa a 2c'\177/k a (1 (8,36)
where ka is defined by Equation 8,19 and c,,. is the cohesion \177nobilized on ab. In the calcula-
tion of ka a reasonable value for 8\" on ab must be assumed. For the passive case the linear
pressure distribution is given by:

= (yz. + + (1 + c w/c') (8.37)

pp q)kp 2c'4k p

where kp is defined by Equation 8.21. In a c' = 0, \177' soil and with the assumption of ,5' = ka
\1773,
a\177id
kp are calculated from Equations 8.14 and 8.17; and the term (yz + q) must be multiplied
by cos\1773. The point of application of the active thrust is determined from
or passive the rele-

vant earth pressure diagrams. Alternatively, a Coulomb wedge analysis may be used to obtain
the active and passive thrusts. The point of application of the active thrust can be determined

using the method of Section 8.4,5 or may be taken as 0.4H above the base, Stability criteria for
both analyses include:
1. N must be locatedwithin the middle third of the base to avoid tensile stress under the heel.
2. The contact pressure at the toe must be equal to or less than/he allowable bearing pressure,
3, The settlement of the toe must be within the tolerable limits,)))
328 Soil Mechanics: Basic Concepts and Engineering Applications

4. The wall systembe safe against

must sliding and overturning.
The long-term stability of a wall may be investigated using the concept of at-rest earth pres-
sure. In this instance, the total thrust on the walt is calculated using ko from Equation 8.4. For
wails with a granular backfill and founded on rock, Duncan et al, (1990) suggestedko = 0.45
= 0.55 for an uncompacted
for a compacted backfill, and ko backfill. Stability criteria were the
same as above but with N permitted to be in the middle half of the base.
Regardless of the linear stress distribution along the line ab, the point of application of the
active thrust may be assumed to be 0.4H above the base. For the walls with clayey soils in the
backfill and foundation, none of the methods above give reliable predictions due to creepof
the material. As a result, the earth pressure charts given by Terzaghi & Peck (1967) are more
reliableas they are based on field experiments. However,the chart solution does not provide a
value for vertical component of the active thrust.

Example 8.7
For the cantilever retaining wall shown in Figure 8.21(a) determine: (a) the soil pressure dis-
tribution along the vertical plane ab, (b) the factors of safety against sliding and overturning
and (c) the maximum and minimum contact pressures under the base. For the backfill soil:
c\" = 0, 0' = 30\302\260,
and \"/= 17 kN/m 3. The angle of friction mobilized under the base is 20\" and
there is no adhesion, The unit weight of the wall material is 24 kN/m 3.

Solution:
(a) Calculate ka from Equation 8.14:
2
-
cos20.0\302\260
- \177/cos 2 20.0
\302\260
_ cos 30.0 \302\260
= 0.441.UsingEquation
ka 8.36:
\302\260 2 \302\260 2 \302\260
cos 20.0 + \177/cos 20.0 -cos 30.0
At z =0, Pa =0. At z =4.5 m, Pa =\302\245zkac\302\260sfj=lT\"Ox4\"5xO'441xc\302\260s20O\302\260=31'7kPa\"

Pa
= 31.7 x 4.5/2 x 1.0= 71.3
kN, and its location is 4.5 13 = 1.5m above the base of the wall.

= = 71.3XCOS20.0 \302\260
= 67.0 kN.
l\177aU,oriaot,tal) =/gab Pa COS\177

= =
Pa sin
= 7 !.3 xsin 20.0\302\260
= 24.4 kN.
Pa(venical) Pat, 13

Calculate the passive force in the front of the wall by assuming \1773'= 0:
= (l+ sin 30.0\302\260)/(t -sin 30.0 =
\302\260)3.0. At z = 0 (in the front of the wall), = O.
kp pp

At z =l.l + 0.4 = 1.5m, =

yzkp
+ 2c'\177Jk\177=
1.5x17.0x3.0 = 76.5 kPa.
pp
= 1.5x76.5/2xl.0 = 57.4 kN and is located 0.5 m above the base of the wall.
Pp

Figure 8.21(b) illustrates the results.

(b) Calculate the total weight W (see Figure 8.21(b) for a definition of W\177to W4):
W\177
= 1.3 x 4.t x 1.0x 17.0 =90.6kN at 1.85 m from the toe.
W2 = 0.4 x 4, I x 1.0 x 24.0 = 39.4 kN at 1.0 m from the toe.
W3 = 0.4 x 2.5 x 1.0x 24.0 = 24.0 kN at 2,5 / 2 = 1.25m from the toe.
Wa
= (2.5
- -
1.3 0.4)t.l x 1.0x t7.0 = 15.0 kN at 0.4 m from the toe.
Thus: W = 90.6 + 24.0+ 15.0= 169.0
+ 39.4 kN.

N =W + vertical component of the active thrust = 169.0 + 24.4 = I93.4 kN,)))

 Lateral Earth Pressure and Retaining Wails 329

71.3 kN
\177 ' 57.4 kN I \177 e
0kN T
T ,__F:_..___-::D__ j _.
\177].7 \177Pa

\177--\177-:\177 0.4 m

(a) (b)
Figure 8.21. Example 8,7.

Shear resistancedue to friction between the base and soil = 193.4 x tan20.0\177= 70.4 kN.
Total force resisting the active thrust = 70,4 + 57.4= 127.8 kN, and the factor of safety against
=
sliding is: Fs= 127.8/ 67,0 1.91. From Equation 8.34:
F v = (57.4x0,5 + 90.6xl.85 +39.4xi.0+24.0xi.25 + 15.0x0.4 + 24.4x 2.5)/67.0x 1,5 = 3.31
(c)Moment equilibrium about the heel:
193.4x = 67.0x1.5+ 90,6(2.5 - 1.85) + 39,4(2.5- 1.0) -
+ 24,0(2.5 1.25)
+15.0(2,5-0.4)- (57.4/1.91)x0.5= 264.96--> x = 1.37,e = 1.37 -1.25 = 0,12m.
Making use of Equations 5.98:

qm,ax
= (193.4/2.5 x t.0)(1.0 + 6x0.12/2.5) = 99.6kPa (at point c, Figure 8.2 l(b)).
qmin
= (193.4/2,5x1.0)(1.0- 6x 0,12 / 2,5) = 55.0kPa (at point b, Figure 8.21(b)),

Example 8.8
A wall 7 m retains a backfill soil with c' = 20 kPa, = 18 \302\260,
and
gravity retaining high \177\"

y = 18 kN/m . A surcharge load of q 30 k.Pa is apphed = to the horizontal upper ground sur-
face. The thickness of the soil in the front of the wall to provide the passive resistance is 3 m
(Figure 8,22).Determine: (a) the soil pressure distribution along the wall, (b) the factors of
safety against sliding and overturning, and (c) the maximum and minimum contact pressures
u\177der the base. Re-solve parts (b) and (c) if the tension cracks are filled with rainwater. As-
sume no base friction and adhesion. The unit weight of the gravity wall is 24 kN/m 3.
Solution:
(a) For a smooth wall =
(1 -sinlS.0\302\260)/(t + sin = 0.528,
18.0 \302\260) = 1/k
a = 1.894,
k\177 kp
At z = 0, Pa = qka
-2c'
\177'\177-a
= 30,0x0,528- 2x20,0x 0\177.528 = -13.2 kPa.
From Equation 8.12, the position of the point of zero pressure is:
zo = 2 x20,0tan(45,0 / 18.0 - 30,0/18,0= 1.39m; the tens lie force above this
\302\260 \302\260
+ 18,0 / 2) point

is ignored. At z = 7.0 m, Pa = (7.0 - 2 x 20.0 x \177-.-.-.\17728= 53.3 kPa,)))

x 18,0+ 30.0)x 0.528
330 Soil Mechanics: Basic Concepts and Engineering Applications

1.0rn

e a-13.2
\\ zo
= 1.39 m

55.0 kPa
d

I
153.4 kN 1

157.3kPa

Figure 8.22, Example 8.8.

Pa
= (7.0-t.39)x53.3t2xl.0 = 149.5 kN and its location is (7,0-1.39)/3 = 1.87m above

the base of the wall. Calculate the passive force in the front of the wall:

[,\177-7\177=55.0kPa.
Atz=0(inthefrontofthewalI),pp=2x20.0x
At z = 3.0 m, = 3.0x lS.0x 1.g94 + 2x20.0x \177 = 157.3 \"kPa.
p\177,
= 3.0x55.0xl.0 =165.0 kN, and is located 1,5 m above the base of tlae wall.
Ppl
= 3.0(157.3- 55.0)/2xl.0 = 153.4 kN, and is located 1,0 m above the base of the wall.
P\177,2

Figure 8.22 illustrates the results.

(b) The weight of the concrete wall has the three components shown in Figure 8.22. Using the

geometry of the watt section we find that WI

= W2 = 96.0 kN, and W3 = 216,0 kN.
The horizontal distances of Wl, W2 and W3 from the heel of the wall are 0,5 m, 1.67 m and 1.5
m respectively, Thus W= 96.0 + 96.0 + 216.0 = 408,0kN.
Total resistance force against active thrust = 165.0 + 153.4 = 318,4k.N,
FS
= 318,4 / 149.5 = 2.13.From Equation 8.34:
Fv = 165.0:,<1.5+ 153.4>, 1,0+96.0\303\2272.5 + 96,0:< 1.33 + 216.0x 1.50/(149.5\303\227 1.87) = 3,91.
(c) Mobilized passiveforces:Pp[ = i65.0/2.13 = 77,5 kN and = 153,4/2.13 = 72.0kN.
Pp2
This ensures horizontal equilibrium of the wall. Moment equilibrium about the heel:
408,0x = 149.5 x1,87 + 96.0\303\2270.5 + 96,0\303\2271,67 + 216.0x 1.50- 77.5\303\2271.5 -72.0\303\227 1.0 = 623.6,
x = 623.6/408.0= 1.528 m, e = 1,528
- 1.50= 0.028m,
Note that the resultant of the contact pressure is located on the left side of the centroid and

thus the maximum contact pressure will be under the toe. Making use of Equation 5.98:
qmax
= (408.0/3,0x 1.0)(1.0 + 6\303\2270,028/3.0) = 143.6 k.Pa (at point c, Figure 8.22),

=
(408.0/3.0\303\2271.0)(1.0-6\303\2270.028/3,0) = 128.4 kPa (at b, Figure 8.22).
qmin point

Pw=forceduetowater=gwxzoxzo/2xl.0=9,81x 1,39x 1.39/2xl.0=9,5kN,

thrust
Total on the back of the wall =Pa + Pu, = 149,5 + 9.5 = 159.0
kN,

FS=318,4/ 159,0=2.0.)))
 Lateral Earth Pressure and Retaining Walls 331

1.5 + 153.4x1.0+ 2.5 + 96.0x 1.33+ 216.0\303\227 1.50

Fv = 165.0x 96.0x
= 3.24.
149.5x 1,87 + 9.5(7.0 - 1.39 x
2/3)
The mobilized passive forces are = 165,0/2.0= 82.5kN and = 153.4/2.0 = 76.7kN.
Pp] Pp2
Similarly: x = 1.640 m, e = 1.64- 1.50 0.14
=
m, qrnax
= 174.1 kPa, and qrmn
= 97.9

8.7 STATIC ANALYSIS OF SH]EET PILE WALLS

8,7.1 Basicconcepts
A cantilever sheet pile wall is a flexural structure and its stability depends entirely on the mo-
bilized passive resistance within the embedment depth. The recommended maximum height
for a cantilever sheet pile is 8 m, Figure 8.23(a) shows a mechanism of failure due to the rota-
tion of the sheet pile about point s. This point is close to the end point and can be determined
by static equilibrium. As a result of the rotation, the state of the stress below point s changes
on both sidesand there is a transition from active to passive at the back and from passive to
active at the front as idealized in Figure 8.23(b). The two unknowns of embedment depth D
and zs. can be found by force and moment equilibrium providing that the mobilized shear
strength parameters on both sides of the sheet pile are known. In a c', soil
\177\" this requires the
iterative solution of two sets of rather complex equations in terms of D and zs. Alternatively
the simplified distribution of Figure 8.23(c) can be adoptedand the stabili.ty can be formulated
to yield the embedment depth, The formulations for D in both c\" = O, d)\" and c,, 0u = 0 soils are
simple and will be discussed in the following sections,
An anchored sheet pile wall is shown in Figure 8.24(a). The recommended maximum

height for an anchored sheet pile is 15 m. The anchor rod is either horizontal or inclined and
its depth from the ground surface is I. m to 2 m. These rods are usually spaced 2 m to 4 m

apart. The active earth pressure in the backfill is resisted by both the passive earth pressure in

q q q

b
/
..... \177( ......
Pas\177,

(a)
T Active
(b)
Passive
(c)
Figure 8.23. Cantilever sheet pile: (a) failure mechanism. (b) idealized lateral earth pressure, (c) simpli-
fied lmeral earth pressure.)))
332 Soil Mechanics: Basic Concepts and Engineering Applications

q q q

Anchor

rod force Active

I
\177c)

',

Active Passive
(a) (b) (c)
Figure 8.24. Anchored sheet pile: (a) failure mechanism, \177) lateral earth pressure distribution for free
earth support method, (c) idealized lateral earth pressure distribution in fixed earth support method.

the front and the anchor rod. This allows a reduction in the embedment or an increase in the

height and, as a consequence, technique

t his is used extensively in excavations and waterfront
constructions. The simple mechanism of Figure 8.24(a) allows rotation about the anchor rod
level without any lateral displacement at this point. Active earth pressure is fully mobilized
along the depth of the backfill. At the front of the sheet pile there is only the passive resistance
(Figure 8.24(b)). As there is no point of contraflexure the equivalent beam analysis, which is
referred to as free earth support, is statically determinate. In dense granular materials the an-
chored sheet pile tends to rotate about point s, which is usually taken as the point where the
bending moment is zero. This rotation causes passive and active states at the back and the
front of the wall respectively similar to the cantilever sheet pile (Figure 8.23(b)). The idealized
earth pressure distribution is shown in Figure 8.24(c). Having the three unknowns of D, Zs and
the anchor rod force, the problem is statically indeterminate unless the position of point s is es-
timated from the equivalent beam method solved by the theory of elasticity (Terzaghi, 1966).
This analysis is referred to as fixed earth support method. If the number of anchors down the

depth exceeds one then the analysis is statically indeterminate and the stiffness properties of
the sheet pile and soil have to be included in the analysis.
The active and passive earth pressures are computed from the general linear earth pressure
distribution expressed by Equations 8.36 and 8.37. It is convenient to combine the lateral earth
pressures of both sides and construct a net pressure diagram. The weight of the sheet pile is
ignored and only the horizontal components of the active and passive thrusts and anchor rod
force are considered. An analysis of both cantilever and anchored sheet piles can be carried
out by assuming that the full active and passive states are mobilized.The computed embed-
ment depth may be increasedby 20% to 40% (this must comply with the standard codes) to
ensuresafety, or alternatively the shear strength parameters or kp
can be reduced. Any change
in the embedment depth will change the lateral stress distribution below the dredge line. The
factor of safety applied to kp in a c', \177' soil must be in the range 1.5 to 2.25 depending on the
magnitude of the effective internal friction angle \177' (Fleming et al., 1992). Lower values of the)))
 Lateral Earth Pressure and Retaining Walls 333

factor of safety correspond to lower values of \177'. For a ca 0u = 0 soil (under the dredge line),
Burland et at. (1981) suggesteda factor of safety of 2 should be applied to cu on both sides
under the dredge line. In both cantilever and anchored sheet piles the application of the factor
of safety to the shear strength parameters or, alternatively, increasing the length of the em-
bedment, will result in the mobilization of only part of the passive resistance. Consequently,
the ideal condition of a linear earth pressure distribution, which applies to a material on the

verge of failure, is no longer valid. This will cause an overestimation of the bending moments
and therefore the maximum bending moment has to be reduced.The traditional correction
method is based on the moment reduction factors suggested by Rowe (1952 and 1957). In this
method, which is based on experimental results, a relationship was developed between the
4
embedment ratio \177= H / (H \303\267D), the anchor rod position, a flexibility number 13
= (H +
D) /
EI and the bending moment reduction factor. It is recommended that the method be usedonly
if a factor of safety is applied to the passive resisting forces. For more flexible walls and lower
values of ko, the results of the traditional methods becomemore reliable. Ports & Fourie (1984
and 1985) applied a numerical analysis to investigate a single anchored retaining wall using
finite elements that obeyed an etasto-plastic constitutive taw. The position of the anchor rod
was selected at the top of the sheet pile and the elastic propertiesthat varied linearly with
depth were known. It was found that a traditional method using the concept of a linear lateral
earth pressure distribution or simple limit equilibrium calculations provide a reliable estimate
of the embedment depth. Although the ko of the soil had little effect in backfill walls, higher
values of ko dominated the behaviour of excavated walls. Lower values of ko correspond to a
reduction in the bending moments and anchor rod force. Figure 8.25 shows the effects of wall
stiffness, expressed in terms of the flexibility nmnber 13, on the bending moments and anchor
rod force for the case where the anchor rod is positioned at the top of the sheet pile. In general
the major shortcomings of traditional analyses are the uncertainty in the displacement behav-
iour, and the distribution of the earth pressure. If the cost can be justified a numerical approach
should be considered.

M / MLimi t
equilibrium (%) e / PLirnit (%)
equilibrium
240 350
300
200
2.0
250
160
200
120
150
80 100
40 50 \"
ko = 0.5
0 0
-1.5 -1 -0.5 0 0.5
Inp In p

(a) (b)

Figure 8.25. Correction of bending moments and anchor rod force (Ports & Fourie, [985),

L)))
334 Soil Mec\302\243anics.- Basic Concepts and Engfneering Applications

4, c'= 0, 0,
\"\177rc',0'
.\" b

(a) (b)

Figure 8.26. Hydrostatic pore pressure distribution in sheet pile walls.

8.7,2 Pore pressure distribution

In the cases when the water table has different levels in the backfill and at the front of the

sheet pile, approximations are made to include the unbalanced force due to this difference to-
gerber with the seepage pressure. In a simplified analysis it is assumed that in the immediate
vicinity of the wall water moves vertically downward in the backfill and upward at the front of
the walt, The hydraulic gradients at both sides are assumed to be equal and constant with
depth, which means equal velocity and seepage pressure on both sides. Alternatively, a flow
net can be constructed to evaluate the pore pressure and seepage pressure distribution. For
equal levels of the water table (Figure 8.26(a)) the active and passive earth pressures are calcu-
lated using the submerged unit weight y' and there is no unbalanced pore pressure,If the soil
under the dredge line is in undrained conditions then the hydrostatic pressures are distributed
according to Figure 8.26(b), The unbalanced net pore pressure diagram is the area of wdeb.
The pore pressuredistribution shown in Figure 8.27(a) is linear on both sides creating equal

hydraulic gradients of i\177and i6 on the right and the left respectively:

iR = (k+d)-u,,/y\177,
\" iL
- uc/Yw-d
h +d d

where ac is pore the pressure at the base and parameters h and d are defined in Figure 8.27(a).
The pore pressure at the base of the wall is found by equating the two equations above:

uc = 2d(h + d)y\177v/(h + 2d) (8.38)

The unbalanced pore pressure diagram is the area of wdc. The net pore pressure at point d is:
u
d
= 2hcig
w/(h + 2d) (8.39)
On the backfill side u,. is lower than the hydrostatic pressure and causes an increase in the sub-
merged unit weight equal to the average seepagepressurealong the wall. At the front of the
sheet pile the submerged unit weight decreases by the same amount, Tl\177e seepage pressure is
defined by Equation 3.10:)))
 Lateral Earth Pressure and Retaining Walls 335

a a

,,,

\342\200\242
. WT
\177Hydrostatic

or c',

(a)

WT

e \"\"

/ \177, d

Hydrostatic
e.<22__
C
\177
\177 MC \1771 \177 MC I

(c)

Figure 8.27. Approximation of po\177 pressure distribution in sheet #lie wails.

J = iYw = (8.40)
\177Yw

The effective unit weights used in the calculation of the lateral ea\177b pressures are \177 for the
backfill and \177L at the front of the wall:

If, at the front of the watt, the water table is above the ground sud\177ce (Figure 8.27(b)), \177he un-

balanced pore pressure diagram is the \177ea of wdec, where:

=
h(d + D)
u\177 y\177,, (8.42)
h+d+D'

Using a similar approach. \177he average seepage pressure within the thickness of h is:

h
J = (8.43)
gu,
h+d+D

L)))
336 SoilMechanics:
Basic Concepts and Engineering Applications

Figure 8.27(c) shows a sectionof a clayey soil with thin layers of a granular material on both
sides of the sheet pile walt, The pore pressureuc is assumed to be equal to the hydrostatic
pressure calculated from the backfill whilst the seepage pressure is appliedonly to the soil at
the front of the sheet pile. The unbalanced porepressureis represented by wdec where u\177 is
the hydrostatic pressure at d and the point e is at the level of the Iower ground surface. This
pressure dissipates linearly at the front of the sheet pile within the embedment depth so that:

J = fr,[w
/ D (8,44)

8.7.3 Cantilever sheetpile in c\" = O, \177O'

soil

A cantiIever sheet pile in a c' = 0, \177' soil is show\177l in Figure 8.28. Under the dredge line the ac-
tive and passive pressures are combined to represent the net pressure diagram. All the active
forces above the zero pressure poi\177t o are represented by their resultant Ra located at a dis-
tance \177 above this point. At point o the active and passive pressures acting from opposite
sides are equal. If there is no water in the vicinity of the wall, the depth of this point from the
dredge Iine Zo is given by:
= -
Zo Pd /'\177(kp k. ) (8.45)
where Pd is the active earth pressure at the level of the dredge line (point b). If water exists on
both sidesof the sheet pile, then:
= - k
Zo P\177I/\302\245'(kp a ) (8.46)
where '{\" -
=
Yw and ka, kp
\302\245\177t are the earth pressure coefficients correspondingto the drained
internal friction angIe. Horizontal equiIibrium of the forces and moment equilibrium about
point c yields the foIlowing equation in terms of z:

Figure 8.28. Net pressure diagram for \177

cantilever sheet pile wall in)))
 Lateral Earth Pressure and Retaining Walls 337

- (2\177C+p o) z =0 (8.47)
\177 ---\177 C2
where Po and C are defined by:

Po = y(H+ = - k
zo)k p -Yzok a, C \"\177(kp a ) (8.48)
and zs is calculated from:

zs - (ppfz -2Ra)/(ppf
+
Ppb) (8.49)
A trial and error method usually provides a rapid solution. The total required length of the pile
is: L = H +D, whereD = zo + z. A simplified solution is obtained if the net passive pressure R
is replaced by a concentrated horizontal force acting at the base of the sheet pile (zs
= 0).

.3
\177
6zoRa
\"Z
6Ra - 0
(8.50)
Pd Pd
Example 8.9
A sheet 6 m of soil with the following
pile supports properties: c' = 0, 0\" = 30% 6' = 17\177,y =
17.3 kN/m 3, kN/m 3. The water
and y'
= 9.5table is on both sides of the sheet pile and is lo-
cated3 m below the upper ground surface. Find the length of the embedment D by applying a
factor of safety of 1.5 to kp at both sides. For the computed embedment depth determine the
location and magnitude of the maximum bending moment.
Solution:
\302\260 = 5.385.
Pot \177' = 30 and =
\1775\" ka = 0.299,
17 \302\260, kp The earth pressure along the wall is inclined
\302\260
at 17 to the horizontal We can introduce the horizontal earth pressure coefficients of kah and

kpj\177providing
that c' = 0: kah
= 0.299\303\227cos17.0\302\260
= 0.286, = 5.385\303\227cos17.0\302\260
= 5.150.
k#\177
Construct the soil pressurediagram above the dredge line and compute zo, Ra, and \177:

dd\" = \"\177Zkah = 17.3 \303\227 0.286

3.0\303\227 = 14.8 kPa.
bb' = dd'+ 9.5\303\2273.0\303\2270.286 = 23.0 kPa. F\177 = (14.8 \303\227
3.0)12 1 \1770= 22.2
\303\227 kN.
=
F 2 14.8\303\2273.0\303\227i.0 44.4 kN. F3
- = =
(23.0-14.8)/2\303\2273.0\303\2271.0 12,3 kN.
The mobilized passivepressurecoefficient = 5.150 / 1.5 = 3.433.
kp,,

, _ Pa _ 23.0
=0,77 m, F4 = (23.0\303\2270.77)/2\303\2271.0 = 8.8 kN.
\177o
C 9.5(3.433- 0\177286)

Ra = F + F2 + + F4 = 22.2 + 44.4 + 12.3+8.8= 87.7kN. Taking moments about point o:

1 F\177

R,\177 \177=
\303\227 (1.0 + 3.0+0.77)
F\177
+ F2(1.5 +0.77) + F3 (1.0 + 0.77) + F4 (0.77 \303\227 2 / 3),

2 = 22 \1772
87.7 \303\227 4.77 + 44.4 \303\227
\303\227 1.77 + 8.8 \303\227
2.27 + 12.3 \303\227 2 / 3 = 232.97,
0.77 \303\227

\177= 232.97/87.7 = 2.66 m. Find po (from Equation 8.48) and necessary coefficients:
= (17.3 \303\227
3.0 + 9.5 \303\227 \303\227
3.433 - 9.5 \303\227
0.77 0.286
\303\227 = 299.0 kPa.
p\177 3.77)

po / C = 299.0/9.5(3.433 - 0.286) = 10.0, 8R / C = 8 \303\227

87.7 /9.5(3.433 - 0.286) = 23.47.
a
2 2
6R
a
/ C = 6\303\227
87.7/[9.5(3.433
-
0.286)[ = 0.5887.)))
338 Soil Mechanics: Basic Concepts and Engineering Applications

2gC + -
Po = 2 x 2.66x9.5(3.4330.286) + 299.0 = 458.05.
2
6Ra(2gC + po)/C = 0.5887 x458.05 = 269.65.

(6RagPo + 4Ra
\177
) / C
2 = x 87.7x
(6 2.66 x 299.0 + 4 x 87.72 )/[9.5(3.433-
0.286)]=
2
502.65.

Substituting into Equation 8.47: z

4 3
+ 10.0z -23.47z - 502,65 0,
2 - 269.65z =

Iteration yields: z = 5.61 m. Thus D = z + Zo = 5.61 + 0.77 = 6.38m, L = 6.00 + 6.38 = 12.38m.
Calculate
Pp and
R:
- 0.286)x5.61
= Cz = 9.5(3.433 = 167.7 kPa, = Po + CZ = 299.0 + 167.7 = 466.7kPa.
ppy Ppb
zs = + -
= (167.7x5.61 2\303\22787.7)/(167.7 + 466.7) = 1.21 m.
(p\177fz -2Ra)/(ppy Ppb)
P ps = C(z- z.,. ) = 9.5(3,433- 0.286)(5.61 - 1.21)= 131.5 kPa.

To find o'c we use the geometry of Figure 8.28 (or 8.29):

%./o'c= (Pps + P
pb)/ P pb, 1.21/o'c =(131.5+466.7)/466.7--\177o'c =0.94 m.
= = 131.5\303\2274.67/2 = 307.0 kN.
Pp pp\177,.(5.61-0.94)/2x1.0
R= 0.94x466.7/2x 1.0= 219.3 kN.

Note that the distance of R from the end point is 0.94 / 3 = 0.31m. Figure 8.29 shows the net
pressure The maximum bending
distribution. moment occurs at a point where shear force is
zero. The net lateral stress below point o is representedby line os and is equal to Cx, where x
is the vertical distance of the poi\177t of interest blow point o.
2 x 1.0 = -Ra + Cx / 2,
2
Thus the shear force SF = -R a + Cxxx[

SF = -87,7+ - 2 / 2 = 0 \177-\177
x = - = 2.42
9.5(3.433 0.286)x \177/2
x 87.7/[9.5(3,433 0\177286)1

The magnitude of bending moment at this point is:

BM =
-87.7(2.66 + 2.42) +[9.5(3,433 -0.286)2.42
2
/ 2](2.42 xl/3) = -374.9 kN.m.
This means that the back of the sheet pile at this point will be in tension, while its front will be
in compression.

P'-a

466.7 kPa
131.5 kPa

Figure 8.29. Example 8.9.)))

 Lateral Earth Pressure and Retaining Walls 339

Example 8.10
Re-work Example 8.9 assu,ning that the passive force R behind the wall acts as a concentrated
force at the bottom of the sheet pile.
Solution:
87.7 / 23.0 z - 6 \303\227
\"\177
Calcul ate z from Equatio,l 8.50: z - 6 \303\227
0.77 \303\227 0.77 / 23.0 = 0,
2.66 \303\227
87.7 \303\227

z
3 - -
17.62 z 46.86= 0, z = 5.17 m. D = 5.17+ 0.77= 5.94 m, take D - 6.0m.
= C: = 9.\302\247(3,433 - 0.286)\303\2275.17 = 154.6 kPa, = 154.6\303\2275.17/2\303\2271.0 = 399.6 kN.
ppf
R= - R a = 399.6 - 87.7- 311.9\"kN. Maximum Ppbending moment remains the same.
Pp

8.7.4 Cantilever sheet pile in cu, =0

\302\242\177u soil

The backfill material is a c', or

preferably
\302\242' c' = 0, \177' soil but the soil below the dredge line is
a purely cohesive soil with =
cu and \177u 0. The net pressure diagram is shown in Figure 8.30.
Noting that in the clay layer ka = = 1, the net
pressure at b becomes 4cu - qd, where qd is the
kp
effective vertical stress at the dredge line level at the back of the sheet pile. As the active and
passive pressure coefficients under the dredge Iine are equal, the effect of the weight disap-
pears and the net pressure below point b remains constant up to the point of rotation. It can be
shown that the net pressure at the end point at the back of the wall is 4c, + qd. For equilibrium
of the horizontal forces and moments about the base:

D(4cu
-
qd )
- Ra
\177 _
\177s (8.51)
4c u

D
2 2Ra D- R\177(12c\177'2+Ra)
-0
-
(8.52)
4c u q\177, (4c,\177 q\177,)(2c\177
+ q\177,)

C I:1
4cu-qd 4Cu+qd
l

Figure 8.30. Net pressure dia\177am for cantilever shec\177 pile in = 0 soil.)))
c,\177,\302\242,,
340 Soil MectuTnics: Basic Concepts and Engineering Applications

For the simplified pressure diagram with Zs = 0, and with R acting as a concentrated force at c:

2 2Ra 2RaT\177
D D- =0 (8.53)
- -
4c\177 qd 4cu qd

The computed D values are increasedby 20% to 40% or, alternatively, the mobilized cohesion
crnu
= cu/F is substituted for cu where F = 1.5 to 2.0. For a stablesheet pile 4c,
- qa'> 0 or:

4cu / F > qa (8.54)

Example 8.11
Re-solve Example 8.9 assuming that the soil under the dredge line is in undrained conditions
with cu = 60 kPa and Ou
= 0. Assume full passive resistance is mobilized.
Solution:
Calculatethe resultant of the active forces abovethe dredge line (see Example 8.9):
R, = F1 + F:Z + F
3
= 22.2 + 44.4+ 12.3 =78.9kN. Distance of Ra from the dredge line:

78.9xg=F\177x4.0+F2xl.5+F3xl.O=22.2x4.0+44.4xl.5+12.3xl.O=167.7, z =2.12 m.

Calculate the necessary coefficients: 4cu -qd = 4x60.0 - (17.3x3.0+9.5x3.0)= 159.6 kPa,

2Ra = 157.8 kN/m, 4cu + qd =320.4kPa.

Ra(12cug + R a) = 78.9(i2x60.0\303\2272.12+78.9) = 632.O.

2cu +qd 60.0
2\303\227 + (17.3\303\2273.0 + 9.5 X3.0)
2 = 0, D = 2.54m.
Substituting into Equation 8.52: 159.6D -157.8D-632.0
Using Equation 8.51, -
z s = (2.54x 159.6 78.9)/(4 \303\227
60.0)
= 1.36 m. From similarity concept:

o'c - 1.36 o'c - 1.36

, o'c = 0.91 m.
4cu+qd (4c,qd)+(4c,--qd)' 320.4 320.4+159.6
= 159.6
Pp
\303\227
[(2.54-1.36) + (2.54-0.91)]/2 \303\2271.0= 224.2 kN.
R = (320.4 \303\2270.91)/2 x l.0 = 145.8 kN. The net pressure distribution along the wall is shown
in Figure 8.31. Apply a 30% increase on D: total D = 2.54 x 1.30 = 3.3m.

a
3
\342\200\242
WT d 14.8 kPa
6.0 m
WT...\177

159.6 kPa b 23.0 kPa

2.12 m
\177'b'

2.54 m --f
1.36 \177 \177 320.4 kPa
R)))

Figure 8.31. Example 8. I I.

 Lateral Earth Pressure and Retaining Walls 341

Example 8.12
Re-work Example 8.11 assuming tllat passive force R behind
the the wall acts as a concen-
trated force at the base of the sheet pile. Increase the computed embedment depth by 30% and
calculate the factor of safety in terms of the mobilized cohesion.
Solution:
Calculatethe necessary coefficients:

2Ra l(4cu -qd) = 2\303\22778.9/159.6 = 0.99,

2Ra\177/(4c . -qd)
= 2\303\22778.9\303\227
2.12/159.6 = 2.10.

Substituting into Equation 8.53: D2- 0.99D-2.10=0,D = 2.03 m.

Calculation of Pp and R:

Pp
= 159.6\303\2272.03\303\2271.0 = 324.0 kN, R = 324.0-78.9 = 2\302\2425.1kN.
D = 2.03 1.3 = 2.65
\303\227 m, L = 6.0+ 2.65 = 8.65 m, For D = 2.65m:

= (4c /F - =
-(17.3\303\2273+9.5\303\2273)]2.65 636.0/F
= -213.1
Pp u q,4)2.65 \303\2271.0 60.0/F
[4\303\227 kN.

Taking moments about point c:

(636.0/F - 213.1)2.65/ 2
- 78.9(2.12 + 2.65) = 0 --\177 F = 1.28.
The mobilized cohesionis 60.0/ 1.28= 46.9kPa.
Pp =2.65(4\303\22746.9 -17.3\303\2273.0-9.5\303\2273.0) = 284.1 kN, R = 284.1-78.9 = 205.2kN.

8.7.5 Anchored sheet pile it\177c', (p\" soil

Most of the sl\177eet piles in this category are constructed from steel and are supported by one or
more anchor rods (or tie rods), which transfer the load to isolated anchorages.A typical net
pressure distribution for a free earth support is shown in Figure 8.32. Summing moments
about the anchor rod and simplifying we obtain:

Figure 8.32, Net pressure diagram for an anchored sheet pile in c'. O' soil.)))
342 Soil Mechanics: Basic Concepts and Engineering Applicatiot\177s

3 2 3Raz\"
-0
z +l.5(Har +Zo)Z (8,55)
- k
\"[(k
p a)
When the water table is at the same level on both sides the effective unit weight and effective
earth pressure coefficientsare used. The ernbedment depth of the pile is D = z + Zo and is
computed using either a vaIue of kp suitably reduced by an appropriate factor of safety or by

using the nonnaI value of kp and simply increasing the resulting D by 20% to 40%. In a fixed
earth support an estimate of the depth of tl\177e point of contraflexure (zero bending moment),
with the distance x measured from the dredge line (positive downwards) based on Table 8.1,
makes the analysis statically determinate (Scott, 1980). Tschebotarioff(1973)proposed that
the point of contrafiexure be taken at the level of the dredge line together with a 33% increase
in the allowable stresses of the sheet pile tnaterial. Williams & Waite (1993) suggested that the
point of contraflexure be taken at the zero pressure point under the dredge line.

Table 8.1. Position of the point of contraflexure in the fixed earth support method.
\302\260 \302\260 \302\260 \302\260 \302\260 \302\260
0\" 15 20 25 30 35 40
x/H 0.37 0.25 t3.15 0,08 0.033 -0.06

Example 8.13

The side of a 10 m excavatiott in sand is to be supported by an anchored sheet piIe walI in

which the anchor rod is 1.3 m below the ground surface. The water table is 2.6 m beIow the
ground surface and there is a uniform vertical surface load of 24 kPa. Soil properties are:
r 3
\177'
= 30 6'
\302\260,= 17 c = = 16.5
\302\260, 0, \302\245 kN/m and 5
/ = 10.4 kN/m 3. Determine:(a) the embedment
depth and the anchor rod force when kp is fulIy mobilized, (b) the embedment depth and the
anchor rod force when a factor of safety of 2 is applied to kp, (c) the location and magnitude of
the maximum bending moment for case (b), (d) re-work case (c) using the fixed earth method

assuming that the point of contraflexure coincides with the point of zero lateral earth pressure.
Solution:
(a) For 0\" ka/, =
= 30 \302\260, 0.286, kp\177,
= 5.150 (similar to Example 8.9).Referring to Figure 8.33:
aa' =qk a = 24.0x0.286
= 6.9 kPa, rid\" = (24.0 + 16.5\303\2272.6)x0.286 = 19.1 kPa.

bb' = (24.0+16.5\303\2272.6 + 10.4x 7.4)\303\2270.286

= 4 I. I kPa.

-o\"
=
Pa/\177'(kp -ka)
= 4I .i/10.4(5.150 -0,286) = 0.81m,
F\177
= 6.9x2.6xl.0 = 17.9 kN, F? -6.9)\303\2272.6/2xi.0= 15.9kN.
= (19.1

F3 =19.1x7.4xl.0 = 141.3kN.
Fa
= (41.1-19.1)\303\2277.4/2xl.0= 81.4kN.
F5 =
(41.1\303\2270.81)/2\303\2271.0
= 16.6 kN, Ra = 273.1 kN.

By taking moments about the dredge Iine we find Ra is 3.68 m above the dredge line. There-
-1.3=
fore z' = 10.0-3.68 5.02 m. Find the necessary coefficientsfor calculation of z:

3Raz' _ 3x273.1x5.02 =81.305.

L5(Har+Zo)=l.5(7.4+l,3+0.81)=\1774.265,
y'(kp -ka) 10.4(5.150-0.286))))
 Lateral Earth Pressure arid Retaining Walls 343

3 2 -
- 0.
Substituting into Equation 8.55: z + l&265z
the coefficients 1.305
\177 Using a trial and
e\177or me\177hod we find: z = 2.22 m. D = z \177zo = 2.22 + 0.81=3,03 \1773 m. Compute co' and F6:
cc\" = y\177(k - k
a)
= 10.4
\177
x 2.22x \177
0.286)
= 112.3 \177a. (5.150
F\177
= (cc'x z)/2xl.0-\177 12.3x2 22/2xl.0 =I24.6\177,
R\177r
+ F 6 = R\177, Rat - 273.[ - 124.6= 148.5\177.
tb) = = 41.1/10.4(5.150/2-0.286) = 1.73 m.
z,\177 Pd/y'(kF -k:\177)

\177 ={41.1xl.73)/2x[.0-35,6kN. R\177 =17.9+15.9+141.3+$1.4+35.6=292.1kN.

T2kmg t\177e moments of F\177 to F5 about t\177e dredge line we find R\177 is 3.39 m above the dredge
line. and therefore z' = 100-3.39- \177.3 - 5.31 m, Fred the necessaD
\177
coefficients:

3Raz' _ 3x292.1x5.31
\177.5(Hur + Z\177,)=l.5(7.4+l.3+1.73)=15.645. =195.465.
y'(kp -ka) 10.4(5,150/2-0.286)
3
195.465 = 0, z = 3.22 m.
2 _
Substituting the coefficients into Equation 8.55: c + 15.6\1775z
D = z + :, = 3.22+ 1.73 4.95 m 5.0 m. Compute co\" and F\177:
= =
- 76.6
co\" = -
\302\245'z(k\177,k,,
) \17710.4 x 3.22 x (5.150/ 2 \1770,286) kPa.
F6 =(cc'xc)12xl.O - 76\1776x3.22/2xl.0 =123,3 \177.
Anchor rod force per metre run of the sheet pile: R,\177r= 292.1 - 123.3 - 168.8

(cl The maximum bending moment occurs at a point where the shear force is zero.
As FI + F2 = 17.9 + 15.9 = 33.8 \177r < R\177r. then this point \177 belong' the water table.
The \177ateral stress below point d is represented by the line d'b' and is equal to 19.1 + \177xk,,
where .r is the vertical distance of the point of interest below d\177Thus the shear force SF is:

SF=Rar-U l -U 2 19.1rxl.O-'[xk, xx/2xl,O-O

SF=168.8-17.9 15.9-19.1x-lO.4xO.286x2/2-O. 1,487.\1772+19.1x-[35,0=0
x- 5.07m < 7.4 m. The magnitude of the bending momen\177 at this point

\177
b , b\" 41.1 kPa
.... -
F ,
5

\177 2.3
\177 kPa c

Fl\177ire 8.33. Example 8.1 \177)))

344 Soil Mechanics: Basic Concepts and Engineering Applications

BM = 168.8(I.3 + 5.07) - 17.9(1.35.07) +

-
15.9(2.6 / 3 + 5.07) - 19,l 5.072
\303\227 / 2

-10.4\303\2270.286\303\2275,073/6 = 556.7 kN.m. This means that the back of the sheet pile at this point
will be in compression while its front will be in tension.
(d) Bending moment at a depth of zo = 1.73m is zero. Thus:
-
Rat (Hat + zo) R a ( z\177 + z ) = O, (8.7 + 1.73)
- 292.1(3.39+ 1.73)= 0, = 143.4 kN,
o Rar Rar

17,9- 15,9
SF = 143.4- - 19.Ix- 10.4 \303\227
0.286x
2
/ 2 = 0, 1.487x+ 19.1x- 109,6 = 0,
2

=
x 4.30m.The magnitude of bending moment at this point is:
BM = 143.4(1.3+ 4.30)- i7.9(1.3 + 4.30) - 15.9(2,6 / 3 + 4.30)- 19.1 4,302/2
\303\227

= 404.6
-10.4\303\2270.286\303\2274.33/6 kN.m,

8.7.6 Anchored sheet pile h\177cu, Ou = 0 soil

In this case the backfill is taken to be a c', \177\"soiI whilst the soil below the dredge Iine is purely
cohesive with cu and \177.
= 0. With the net pressure diagram shown in Figure 8.34 and full mo-
biIization of c., the sum of the moments about the anchor rod level yields:

2R a z'
D2 + 2Har D =0 (8.56)
4c, - qd
Equilibrium of the horizontal forces is used to determine the force in the anchor rod, The gen-
eral expressionfor the mobilized passive thrust is:
4cu
=
Pp --qd ) (8.57)
D<---\177

where F is the factor applied to

of safety Cu. If the water levels on both sides are different the
area wdeb of Figure 8.26(b) must be taken into account in the force and moment equilibrium
equations,
q

Har

(b' \302\242 b

D
=
PP Cu, \177u 0
t\302\242.._
c'

Figure 8,34. Net pressure diagram for an anchored sheet pile in c,,, Ou = 0 soil.)))
 Lateral Earth Pressure and Retaining Walls 345

Example 8.14
In an anchored sheet pile the backfill has the same profile as in ExampIe 8,13. The soil below
the dredge line has cu = 75 k.Pa and d?u
= 0. Determine: (a) the embedment depth and the an-
chor rod force when a factor of safety of 2 is applied to cu, (b) the factor of safety when there
is a 3 m sudden drop of the water table at the front of the sheet pile,
Solution:
(a) R a = 17.9 + 15,9+ 141.3
+ 81.4 = 256.5 kN.

Taking moments about the dredge line we find Ra is 3.94 m above the dredge line. Thus
z' = 10.0-3.94-1.3 = 4.76m.
qd
= 2.6\303\227
16.5 + 10.4
7.4\303\227 = 1 i9.9 kN. Use Equation 8.56:

D
2
+ 2\303\2278.7D -2\303\227256.5 \303\2274.76/(4\303\22775.0/2 -119.9)
= 0, D 2 + 17.4D-81.12= 0, D = 3.82 m.
=l
Pp =3.82(4\303\22775.0/2-119.9)
l5.0 kN,
Rar -Pp =256.5-115.0=141.5kN.
= R
a
(b) CaIculatethe magnitudes of the horizontal forces due to backfill water and the water at the
frontand their moments about the level of the anchor rod:

= 2 12 = 9.81\303\227 = 268.6 kN.

Yw H
Pa.'(l\177ack\177
7-4212
w(back)
\303\227
1.0

M w\177 = 268.6(8.7 - 7.4/ 3) = 1674, 3 kN. m.

back)
= 2 /2 = 9.8 (7.4-3.0)2/2\303\2271.0 = 95.0 \177kN.
Pw(from) \"[wHw(front)
M =
95,0{8.7
- (7.4 - 3.0)/ 3] = 687.2 kiN. m.
\177'(from)

Sum moments about anchor rod: Raz = + D/ 2)

the
Mw(back ) -Mw(from ) Pp( Har
256.5\303\2274.76+1674.3 -687.2 =
Pp (8.7 +3.82/2) --r Pp = 208.1 kN.

Pp
= D(4c,, / F -qj ) = 208.1,3.82(4\303\22775.0/F 119.9) = 208.1 --r F = 1,72.
-

The new anchor rod force is: = 256.5 - 208, \177

+ 268.6- 95.0 = 222.0 (a 57% increase).
Rat

8.7.7 Earth pressure on sheet piles in an excavation

A general arrangement of a braced cut is shown in Figure 8.35(a) in which horizontal struts
support the sheet pile system. A traditional pressure pro-
approach is to use the apparent earth
files of Figures 8.35(b) to 8.35(d) that are & Peck (1967),Peck
based on the work of Terzaghi
(1969)and Flaate & Peck (1973). These diagrams do not represent the actuaI lateral earth
pressure but are an envelope to the approximate lateral stress distribution due to the force in
each strut and is constructed from the measurement of the strut forces. For sands the uniform
lateraI pressure of Figure 8.35(b) is used,where ka is calculated from Rankine's theory. For a

purely cohesive soiI a trapezoidaldistribution is used that depends on tile magnitude of the
stability number defined by n = yH / cu\177For n > 4 the lateral distribution of l\177igure 8.35(c) is
more appropriate, where the parameter m varies from 0.4 to \177.0. with lower values of m for
soft clays. For n < 4 the clay soiI is in elastic equilibrium and it is convenient to use the distri-
bution of Figure 8.35(d). As the stability number approaches 8, pIastic equiIibrium dominates

the backfill and the soiI unde\177leath and the heave of the bottom of the excavation will Iead to)))
346 Soil Mechanics: Basic Concepts and Engineering Applications

-,,\177-..........\177,,
Sheet
pi\177e
H
Strut
\177

0.25H

(a)

0.65K a '(H ka 0.2-0.4'(H

K
(b) a=l-4m/(YH/cu) (d)

Figure 8.35. Apparent earth pressure profiles on sheet piles in an excavation.

a shear failure of the system. A slightly different profile of the apparent pressure diagram in
sand and clay may be found in Tschebotarioff (i973). Note that the traditional method does
not account for the effective stress analysis and multi-layer soils. Bowles (1996) suggestedthe
use of an average of the two pressure diagrams obtained from Rankine's theory and the at-rest
condition. Alternatively, either the active earth pressure coefficientor the magnitudes of the
lateral stresses may be increased by a factor in the range 1.0 to 1.3.
The analysis of the sheet pile is carriedout using the following methods:

1. Calculatethe force strut by assuming

at each that the area of the pressure diagram between
the two struts is equally shared by the adjacent struts. Note that this satisfies equilibrium of the
horizontal forces, but moment equilibrium is not satisfied.
2. Assume the top and the bottom struts have fixed connections with the sheet pile, while the
other have hinge connections. Use the above mentioned
struts pressure diagrams to analyse the
sheet pile beam. The probIem is statically determinate,
3. Using the average pressurediagrams obtah\177ed from ka and ko as the actu,al response of the
backfiI! soil, treat the sheet pile as a continuous beam. This probIem is statically indeterminate.

In a slurry excavation
the force P due to hydrostatic pressure from the slurry material made
of water and bentonite
(montmorillonite group) supports the cast in-situ concrete sheet pile.
Generally bentonite can be used in combination with soil and other additives (e.g. cement) to
improve the flexibility and decrease the permeability. During the excavation a thin imperme-
able layer will be formed within the slurry at the face of the excavation. This layer prevents
the loss of slurry into the soil and allows the construction of the concrete elements to cover the
\302\260
face of the excavated soil. A Coulomb wedge analysis can be adoptedwith ct = 45 + 0' / 2 to

determine the maximum value of P provided by the slurry. This allows the unit weight of the

slurry to be obtained and thereby the proportion of bentonite that shouId be used.)))
 Lateral Earth Pressure and Retaining Walls 347

8.7.8 Common types of plate, concrete, and ground anchors

Common types of sheet pile anchorages are shown in Figure 8.36. The vertical plate anchor
shown in Figure 8.36(a) may be made of steel plates, precast or in-situ concrete. The anchor
(or tie) rods are spaced (preferably) equidistantly and are either horizontal or slightly below
the horizontal. Each rod is secured to a plate anchor that may be astrip (thereby facilitating
plane strain conditions) or a plate of finite dimensions. Its holding capacity is mainly derived
from the passive resistance at the front of the plate, which is reduced by the active thrust cre-
ated at the back of the plate. The ultimate value of holding
capacity is determined by introduc-

ing suitable failure mechanisms basedon the development of a passive failure surface from the
tip of the plate. This type of failure mechanism (general or local) has been verified by labora-
tory tests on models (Hueckel, 1957; Das, 1990). In a general failure meclmnism the ratio of
the depth to the height of the plate is small and the failure surface intersects the ground sur-
face. In a localized failure a cylinder of soil with a radius equal to the height of the plate is
considered. The length of the anchor rod must exceed the active zone at the back of the sheet
pile and the passive zone at the front of the plate anchor.
A block concrete (or dead man concrete) anchor is shown in Figure 8.36(b). Its holding ca-

pacity relies mostly on the passive resistance at the front of the block. The weight of the block
and the soil resting on its upper surface mobilizes shear resistance on the upper and lower con-
tact surfaces, further increasing its holding capacity. If, for a specified geometry, the analysis
shows a possibility of sliding or overturning, then the block couldbe supported by a group of
piles as shown in which the piles at the front will be in compression while those at the back
will be in tension.
A groundanchor consists of an anchor rod grouted in a drilled hole as shown in Figure

8.36(c). Typically, these drilled holes have diameters between 100 mm and 300 mm over a
length of some 10m to 30 m. A ground anchor depends entirely upon the shear resistance over
its grouted length. This resistance is a function of the method of construction, the pressure by
which the grout is inserted into the hole, the average effective vertical stress in the vicinity of
the grouted portion and the dimensions of the anchor. In Europe a Coulomb wedge-type analy-
sis is normally used to evaluate the ultimate holding capacity of a ground anchor (e.g. BS
808i, 1989). This method allows the sheet pile and the anchor to be analysed as a unified sys-
\302\260 \302\260
tem. Anchor rods are often inclined at I5 to 25 below the horizontal to increase the effective
vertical stress and thereby the corresponding shear resistance. Steep slopes,up to 2 vertical to

// t
\177 -.
\"-
IN..
I I\177la[e .
,
/\342\200\242 \"--
- \177
.\177,\"\177I
/\177 ,
t Ancnorgrout

I, I I

I i # \177\"
I

(a) (b)

Figure 8,36. Common types of sheet pile anchorages.

L)))
348 SoilMechanics: Basic Concepts and Engineering Applications

1 horizontal, can be used to provide sufficient effective vertical stress, facilitate gravity grout-

ing and possibly avoid underground utilities.

The anchor rod must meet the requirements of the relevant standards. Generally anchor
rods are made from hot-rolled alloy steel bars strengthened by cold working that produces a
high tensile Anchor rods may be housedin pipes to prevent possible damage caused
strength.
by backfill or ground heave, and must be covered by a material specified in the
settlement
relevant code in order to prevent contamination and corrosion by minerals and domestic waste.
The factor of safety is defined as the ratio of the ultimate holding capacity to the anchor rod
force calculatedfrom equilibrium of the sheet pile. The factor of safety is normally in the
range 1.5 to 2.5 depending on the type of the anchor, the soil and the design life span of the
system. For temporary anchorages a factor of safety of 1.5 can be used but for permanent an-
chors higher values of the factor of safety should be adopted, In ground anchors where the an-
chor has a steep slope (e.g. 2 vertical, 1 horizontal) the factor of safety must be increased
(from 1.5 for horizontal) to 2.

8.7.9 Holding capacity of plate anchors

The design of a plate upon an understanding of its behaviour,
anchor relies the shear strength
properties of the soil magnitudes of and the
the angles of friction mobilized on both sides of its
surfaces. It is normal practice to assume that the plate is rigid and undergoes sufficient transla-
tion towards the retaining structure to create both the passive and active states. Figure 8.37 il-
lustrates the vertical section of a typical plate anchor system in sand. Based on the design
methods suggestedby Teng (1962), Ovesen & Stromann (1972), Tschebotarioff (1973) and
Meyerhof (1973), a number of approaches are available. In Teng's method the active and pas-
sive thrusts are calculated using Rankine's earth pressure coefficients. If the ratio B / h < 1 / 3
to I / 2, the ultimate holding capacity in plane strain conditions is obtained from equilibrium
of the horizontal forces (Das, 1990):
R = /2 - / 2 - qk h) / cos r I
u (\177th
2kp \177h2ka a (8,58)

\177\"..,C b.,\" h - B
\"\" Zone 1
... Zone 2

,,\"
f ep-\"\177 ,,
\",.,.,.

Figure 8.37. Ultimate holding capacity of a vertical plate anchor.)))

 Lateral Earth Pressure and Retaining Walls 349

For safety reasons, the effect of any surface loading is usually ignored in the calculation of the
passive thrust. For B / h values greater than 1 / 2 the lateral soil pressure is estimated according
to that given previously in Section 8.3. For a plate of finite width the effect of the shear resis-
tance developed on the vertical faces of the failure surface (plane strain in the previous case) is
taken into account, thus:

COS r
\177l.

where ko is the coefficient of earth pressure at rest that may be assumed to be 0.4 and L is the

width of the plate. In this analysis vertical equilibrium may not be satisfied.
In the original & Stromann (1972) method, equilibrium
Ovesen of the forces both horizon-

tally and
vertically is considered in conjunction with the following assumptions:
1. The anchor rod is horizontal.
2. The passive and active failure surfaces are combinations of planes and spiral surfaces and
reflect the actual failure mechanism.
3. The mobilized friction angle on the pIate on the active side (zone 2, Figure 8.37) is equal to

4. The mobilized friction angle on the plate on the passive side (zone 1,Figure 8.37) is 6' < 0\"
and is computed from equilibrium of vertical forces. The weight of the plate is ignored.
5. Moment equilibrium is not satisfied.
For an inclined anchor rod the following analysis takes account of equilibrium in both hori-
zontal and vertical directions. For horizontal equilibrium:

+ cosS'= 0, = cosr'-
Ru c\302\260srl Pa Cos0\"- Pp R
u (Pp Pa cosO')/c\302\260Srl ' thus:

1
- (8.60)
Ru= -
[(yh2k\177 ; cos6')/2 (yh2ka cos,')/2]
COS \177

For vertical equilibrium, neglecting the weight of the vertical plate,

Ru =
[(\177'h ka sin0/2)
- sink\342\200\242/2)]/sinr I (8.61)
(yh2kp

Equating Equations 8.60 and 8.61, \1775'may be obtained by a trial and error method. By substi-
tuting 5' into Equations 8.60 or 8.61, the ultimate holding capacity can be calculated. For a
horizontal anchor rod the fight-hand term of Equation 8.61 is zero: k a sin6' = 0
sin\177'-kp

ka = (8.62)
kp sinr'/sin\177'

Substituting ka and kp from Equations 8.20 and 8.22 we find:

'
sinr' = 9
[(\177-co-\1777-4sin(\302\242'+ 6')sm\177' ')/(.,/-\177-\177+' 4sin_0\" 's,n0 ,)]2 sm\177\" (8.63)

The mobilized friction angle 8' is found by a trial and error method. Substituting Equation
8.62 into Equation 8.60 the ultimate holding capacity R, may be expressedas:
2 -
R,,
=
7h k
p
sin(\302\242\302\242 8') / 2 sin \302\242\177' (8.64))))
350 Soil Mechanics: Basic Conceptsand Engineering Applications

For a horizontal anchor rod in a soil in an undrained condition, a breakout factor in dimen-
sionless form Fe is defined by Tschebotariff (1973) to give the ultimate holding capacity as:

R,,
= (8.65)
Fcc,,h

According to laboratory smaIl-scale model tests, the magnitude of Fc increases with the depth
ratio h/B and remains constant after a critical depth ratio is achieved. This latter ratio defines
a shallow anchorage condition. The following empirical equations were proposed by Meyer-
hof (1973):

Fc = 1.2bib square anchor, critical bib = 7.5 (8.66)

Fc = hi B strip anchor, critical h/B = 8.0 (8.67)
Further tests showed that the critical depth ratio depends upon the magnitude of the undrained

cohesion; for square anchors this ratio is approximately 7.4 (Das et al., 1985).It has aIso been
suggested that Equations 8,66 and 8.67 underestimate the actual ultimate holding capacity
(Das, t990). The breakout factor can also be determined using a suitable upper bound failure
mechanism. Ignoring the active failure at the back of the anchor pIate, a plane strain mecha-
nism, similar to the tunnel heading mechanism of Figure 7.26, can be applied. With sufficient

accuracy Equation 7.61, suggested by the author for a plane strain tunnel heading, can be rear-
ranged to represent the breakout factors for depth ratios up to i0:

Fc =2.64+21n(h/B), (lto2), Fc =4.0+21n[(h-B)/B], (2 tol0) (8.68)

It is conunon to appIy a factor of safety of 3 to the ultimate hoIding capacity. In the analysis of
a sheet pile system, the anchor rod force is calculated by using a factor of safety against the
passive state at the front of the sheet pile, thus for the plate anchor a reduced factor of safety of
1.5 to 2 can be used.

Example 8.15

the ultimate holdin\177 capacity of a vertical anchor plate in sand using the following
Calcul,ate =
data: d) 35 y
= 17.5
\302\260, kN/m\177 B = 2 m, L = 2.5m, and h = 2 m.

Solution:
From Equations 8.6 and 8.9:
k = tan2(45.0 \302\260-
35.0\302\260/2)
= 0.271, = tan2(45.0\302\260 + 35.0\302\260/2)= 3.690.
a kp
Pa
=
yh2ka/2 = 17.5x 2.02x0.27t/2= 9,5kN/m.
= = 17.5x2.02 x3.690/2 = 129.1 kN/m.
Pp yh\177kp

yh3k + = 17.5 x 2.03x 0.4(3.\177'\177 + 0.-]-d\177)/3 = 45.6 kN. From Equation 8.59:
o (k\177p k\177a)/3
Ru = - ) +
= 2.5(129.1-9.5) + 45.6= 344.6 kN.
L(Pp Pa Pf

ExampIe 8.16

Re-work Example 8.15 assuming that at the active side, the angle of friction between the soil
and the vertical plate is mobilized to)))
 Lateral Earth Pressure and Retaining Walls 351

Solution:
\302\260
= 35\302\260 \177'- ']sin(35'0\302\260 6\") sin
\303\267 35,0
Substituting 0' into Equation 8.63: sin 8' = \302\260
sin\177',
\302\260
\177 + \177/sin 70.0 sin 35.0 \"]2

Using a trial and error procedure5' is found to be 2.07 \302\260.

From Equations 8.22:
= - \302\260 \302\260
kp [cos35,0\302\260/(\177
sin 35.0
+ 2,07 \302\260) )]2 = 3.96.
ffsin(35,0
Using Equation 8\17764:

= (17,5x2.02 \302\260
R \303\2273.96\303\2272,5/2)sin(35.0\302\260- 35,0 = 328.4 kN.
u 2.07\302\260)/sin

8.7. t0 Holding capacity of block concrete anchors

An idealized free body diagram of a blockconcrete anchor is shown in Figure 8.38. The factor
of safety may be applied to either Ru or
kp. Equilibrium of horizontal and vertical forces al-
lows the unknown magnitudes of Ru and N2 to be computed whilst the point of application of
the base reaction N2 is found by moment equilibrium.

Example 8.17
For a strip concrete block anchor the following data are kqown: B\" = 1 .n,/3 = 1.3 m, t\177= 2,5
m, q = 30 kPa, \177'= 18 kN[m 3, \177'c= \177[conerete
= 23.5 k_N/m 3, c' = 0, \177'
= 35 \302\260,
and 6\" = 17 \302\260.
Deter-
mine the ultimate holding capacity of the anchor assuming a horizontal anchor rod and a factor
of safety of 2 for the mobilized kp,

Solution:

From Equations 8.20 and 8.22: ka= 0.256 kp = 6.767, and therefore (as c' = 0):
\302\260
= 0.256\303\227cos17,0\302\260
= 0\177245 and 6,767\303\227cos17.0= 6.471. =
kah kpi t
a'a = (30.0 + t8.0\303\227t,2) \303\2270.245 = 12.6 kPa, b'b = (30,0 18.0\303\2272.5)\303\2270.245 = 18.4 + kPa,
= + F = 12.6\303\2271.3\303\2271\1770+(18.4-12.6)\303\2271,3\303\2271.0/2 = 16.4+3.8 = 20.2 kN.
P,t, F\177 2

v
Ppv

\177
I T N\177 I

Figure 8.38. Ultimate holding capacity of a concrete block anchor.)))

352 Soil Mechanics: Basic Concepts and Engineering Applications

za
=
(F\177zt
+
F2z2)/Pa\177,
= (16.4xl.3/2 + 3.8xl.3/3)/20.2 = 0.61m.
d\177/=- 18.0xl.2x6.471 = 139.8 = 291.2
kPa, e'e = 18.0x2.5x6.471 kPa.
=
F3 + F,\177
= 139.Sxl.3xl.0 + (291.2 - I39.8)x1.3xl.0/2= 181.7 + 98.4 --= 280.1 kN.
Pph
= (F3z +
F4z 4) /
= (181.7x 1.3/2+ 98.4 \303\2271.3/3)/280.1 = 0.57 m.
zp 3 ?pt,
N
t
=
B'xl.O(yh + q) = 1.0x 1.0(18.0xl.2+ 30.0)
= 51.6 kN.
Pav
=
Pal, X tan 6\" = 20.2 x tan 17.0 \302\260
= 6.2 kN, = x tan 6' = 280.1x tan 17.0
\302\260
= 85.6 kN.
Ppv Pph
W = B'xB\303\227l.0\303\227yc
= 1.0xl,3x 1.0x23.5 = 30.5 kN. From vertical equilibrium:
N2 =N 1
+ W + Pay
-
Ppv
= 51.6 + 30.5+ 6.2- 85.6= 2.7 kN.

T1 = N tan6'=51.6xtanlT.0
I
\177=
15.SkN, =
T2 N 2 tanG'= 2.Txtan17.0
\302\260
=0.8 kN.

For horizontal equilibrium:

= - = 280.1 - 20.2+ ! 5.8 + 0.8 = 276.5kN.
Ru Pah + Ta + T2
Pp\177,

Applying a factor of safety of 2 for the mobilized

6.2- 85.6/245.5
kp:

N2m
= N + W +
I Pay
- [ 2 = 51.6 + 30.5+ = kN.
Ppv
= N tan \302\260
2 8' = 45.5 \303\227
tan 17 = 13.9 kN.
T2m
Rat = Pphm
-
Pah + T1 + T2m = 280.1 / 2- 20.2+ 15.8 + 13.9 = 149.6 kN.
Factor of safety against holding capacity
= 276.5 / 149.6 = 1.85. Assume the anchor rod is
0.75 m above the base and check that the force N?m is located within the base (0 < x < 1.0m):
x = [ Pah x za + (W + N 1 )B' / 2 + Rar x za\177
- x B
T\177
-
Pphm
X Z
p
-
Ppvm
x B'I / N2m, x = 0.49 m.

8.7.11 Holding capacity of ground anchors

Based on the soil-grout ultimate bond stress '\177ult, the length of the grouted zone lr is given by:
Ru x
- F sRar
=
tr (8.69)
%011 \"Cult %0tl \177ult

where the factor of safety,

F is s is the ground anchor spacing, Rat is the anchor rod force and

D\177 is the of the hole. The minimum grout length

diameter normally lies between 3 m to 4.5 m
(Littlejohn & Bruce, 1977; Munfakh et aI., 1987) and must comply with the relevant code. The
average ultimate bond stress depends on the type of the soil and the grout pressure (PTI,
1996). In cohesive soils, this bond stress is between 30 kPa to 380 kPa for high grout pressures
and between 30 kPa to 70 kPa for low grout pressures. In cohesionless soils the ultimate bond
stress for a low pressure grout is 70 kPa to 140 kPa and for high grout pressure is 80 kPa to

1380 kPa. For soft and hard rocks a low grout pressure is used and the ultimate bond stress
varies from 150 kPa to 3100 kPa with 1400 kPa as a maximum value for soft rocks. The aver-
age ultimate bond stress can be obtained by in-situ testing that includes pullout, performance
and proof tests. In the latter two tests the ground anchor is loaded up to between 1.3 and 1.5
times of the design load. Cheney (1988)recommended test procedures and the requirements to
be used in the selection of the anchors. Further information about design considerations, meth-

ods and testing procedures may be found in Xanthakos (1991), Xanthakos et al. (1994) and)))
 Lateral Earth Pressure and Retaining Walls 353

Schaefer et al. (1997).Alternatively, the following equation can be used to estimate the aver-
age ultimate bond stress:

\"\177ult
--
Ozm k tan6 + ca (8.70)

where o-'\177 is the effective vertical stress at the midpoint of the grouted length, k is the lateral
earth pressure coefficient between ka and ko, 5\" <-\177' is the fr\177ction angle mobilized on the
surface of the grout cylinder, and c a < c' is the cohesion mobilized on the soil-grout surface.

8.7.12 Numerical analysis of sheet pile walls

Sheet pile walls can be analysed using the beam-foundation model discussed previously in

Chapter 5 by assuming a suitable embedment depth. Lateral earth pressure is applied to the
vertical beam only the zero pressure point. The anchor rod and the
above soil resistance below
the zero pressure point are idealized by a set of springs. The spring stiffness for the anchor rod
(per unit width of sheet pile) is given by:

- AEcos\177
Kar (8.71)
sL
where E is the Modulus of Elasticity of the anchor rod material, A is the cross-sectional area, L
is the length of the anchor rod, s is the anchor rod spacing and \"q is the slope of the anchor rod
with horizontal. For a unit width of soil, the spring stiffness under the dredge line is:

K i =
ai' + a i'-' 1
ksi (8.72)
2

where ai andai-1 are the distances to the springs adjacent to spring i and ksi is the modulus of

subgrade reactionat spring i. Bowles (1996) suggested the use of the following equations for
the estimation of modulus of subgrade reaction at a depth z:
\177
k
s
= A
s + Bsz n > 0 (8.73)
3
k s = 40qu qu in kPa, ks in \"kN/m (8.74)

where As, Bs and n are coefficients and z is the depth of the point. The term qu in Equation
8.74 represents the ultimate bearing capacity of the ground at a certain depth that can be esti-
mated theoretically if the shear strength parameters c' and \177' are known (Chapter 10) - it may
also be obtained from field experiments. Using Equation 8174 at several specified depths al-
lows unknown parameters of As, Bs and n to be obtained. If xi is the horizontal deformation
th\177
of the i spring, the force in the spring is:
= \303\227
Ri Ki xi

The solution of the pile beam can be arranged

equivalent sheet in terms of the unknown xi val-
ues. By dividing the
corresponding spring force by the effective area of 1.0 \303\227
(ai + ai - t) / 2,
the average normal stress acting on the vertical interface can be evaluated. If any of these
normal stresses or the spring deformations are higher than specified allowable values, the em-
bedment depth has to be increased.)))
354 Soil Mechanics: Basic Concepts and Engineering Appticatiot\177.s

8.8 INTERNALLY STABILIZED EARTH RETAINING WALL

8.8.1 Geosynthetic reinforcement

Factory manufactured geosynthetic products are gradually replacing conventional reinforce-

ment materials such as meta/ strips, metaI bars and weIded wire grids as soil reinforcement
materials. Major applicationsof geosynthetics in ground improvement, reinforcement and
treatment have been given by Koerner (1994) and incIude earth walls, steep slopes, clay fitl
slopes, embankments on soft soils, pile embankments, unpaved road stabilization, and treat-
ment of areas prone to subsidence. In retaining walls economic benefits and a reduced con-
struction time may be realized by using a poor quality soil reinforced by geosynthetics as
backfill. In comparison with reinforced concrete walls, a 50% reduction in cost is predicted for
a 5 m high wall; for a 10 m high wall the reduction in cost could be as much as 250%.

Geosynthetics are manufactured polymeric materials that are currently available in the
following forms (Schaefer et aI., 1997):

1. Geotextiles (GT); permeablepolymeric materials comprised of textile yarns used for filtra-
tion, drainage, separation or reinforcement.
2. Geogrids(GG);grid-like sets of interconnected polymer ribs used for reinforcement.
3. Geonets (GN); net-like sets of interconnected polymer ribs used for the transmission of liq-
uids within the plane of their structure.
4, Geomembranes (GM); impermeable barriers used to contain various solids and liquids.
5. Geosynthetic clay liners (GCL); factory, manufactured hydraulic barriers consisting of a thin
layer of bentonite supported by geotextiles and/or geomembranes.
6. Geocomposites (GC); a combination of any of the above geosynthetics that sometimes in-
cludes natural soils.

Since the basic objective of the reinforcement in the backfill is to transfer the active thrust to

the reinforcement elements, the working tensile strength of the element must be established
accurately. Due to the high extensibility of the most geosyntheticsthey are subject to instalIa-
tion damage and creep during the design life. Construction damage may Iower the tensile

strength by as much as 30% which is equivalent to a factor of safety of Fc = 1.3 to 1.4 on the
initial (nominal) tensile strength. The magnitude of Fc may vary between 1.1 and 1.6 (Task
force27, 1990)and its minimum value must comply with the relevant design code.
In general, the load-extension behaviour is a function of time. Laboratory results are com-
monly presented in two sets of curves showing: (1) the relationship between time and uItimate
tensiIe strength; (2) the
relationship between extension
strength and tensile
specified time for

intervals. The time-tensilestrength relationship is approximated by a line if time is pIotted on

a logarithmic scale and is constructed for the time period over which the creep data is avail-
able (100013 hours according to ASTM D-5262, 1997). As time increases, the ultimate tensile
strength decreases. To find the ultimate tensiIe strength for the required design life (say t20
years), extrapolation of the linear relationship is permitted providing that a reduction factor of
about 70% is applied to the computed tensile strength. This is equivalent to a creep reduction)))
 Lateral Earth Pressure and Retaining Walls 355

factor of CRF = 40% (higher for woven geotextiles) applied to the nominated ultimate tensile
strength. In the absence of any creep tests, the default values of the corresponding code must
be applied. In addition, a general factor of safety of approximately Fs = 1.2 (slightly higher for
retaining walls) is applied to take account of the possibility of reaching a limit state and other
uncertainties during the design life (Task force 27, 1990).If necessary, factors of safety
against chemical damages and biological degradations (F\177) must be considered. Consequently,
the allowable tensile strength may be computed from the following equation:
x CRF
- Tult
Tal I (8.75)
F\177.x.\177 x\177
where Tazl and Tult represent the allowable and ultimate tensile strengths respectively. The cor-
responding extension (tensile strain) to the design load determined from the
can be set ofthe
load-extension-time relationship constructed in the laboratory. For specifiedtime intervals
such as one hour, one month, one year and 5 to 10 years, the stress-strain type curves are de-
termined by conducting tensile tests on the geotextile materials. Extrapolation is used to esti-
mate the extension during the design life; however, with the above factors of safety consid-

ered, the initial and final extensions become nearly identical because the time isochrones share
a common slope up to 40% of the ultimate tensile strength (Exxon, 1992). Thetensile force in
the reinforcement is redistributed in the soil through the surface area of the reinforcement.
Laboratory tests are used to determine the bond stress between the soil and the reinforcement.
The friction angle mobilized between the geosynthetic reinforcement and the adjacent soil 6b
may be related to the internal friction angle of soil \302\242'
using the following equation:

fz, tan
tan = (8.76)
6/\177 q\177'

The magnitudeofft, depends on the type of the reinforcement material and varies from 0.5 to
1.0. It is determined from the recommended standard laboratory tests such as a directshear test
or a full-scale pullout test. In the direct shear test, the reinforcement material is placed on the
enforced failure plane. In a pullout test the reinforcement material is pulled through the soil

along the separation of the two half boxes of the shear box. For granular soils, the apparent
friction angle from pullout test is lower than the results obtained from the direct shear test.

8.8.2 Reinforced earth retaining walls

Depending on the type of the reinforcement, two design methods are available. For a rein-
forcement with high extensibility such as geotextiles, the active state in the earth wall is fully
mobilized and the active Rankine earth pressure coefficientka is used. In a granular backfill
with c' = 0 and of the potential
0', the position failure plane may be computed using the Cou-
lomb wedge method. For a vertical wall, horizontal ground surface and with = 0, the failure
\1775'

plane makes an angle of a = 45 \302\260

+ 0' / 2 with the horizontal. With a sloping ground surface

and \1775'= [3, the position of the failure plane may be formulated from the Coulomb wedge
analysis. Using the same terms given in Equation 8.19, the magnitude of c\177is given by:

- = [- - / B
tan(o\177 \177') tan(\177\" [3) + \177-\177-] (8.77))))
356 Soil Mechanics: Basic Concepts and Engineering Applications

A = tan(0\"- 18)[tan(*'-[\177) + cot(,'+ O- 90\302\260)][1 + tan(5' + 90 \302\260

-
0)cot(,\" + O -90\302\260)]

\302\260 -
B = 1+ tan(8' + 90 - O)[tan(\177'- I\177)+ cot(*' + O 90\302\260)]

In the absence of hydrostatic water soil, the axial tensile load transferred to each hori-
in the
zontal reinforcement element (Figure 8.39) can be calculatedas:
= \303\227 -- (8.78)
Pr Pah sx \303\227sz (ka\177Tz cOS\177)SxSz

where Pr is the tensile force in thestress at depth z within the active

element, \302\242\177z
is the vertical
zone, Sx and sz are the
horizontal spacing the of the and vertical
reinforcement elements re-
spectively. The embedment length lr is calculated to resist the tensile force pr. For a strip ele-
ment of width w, the maximum tensile toad pt due to the frictional bond on both sides is:

= = (8.79)
Pt 2crzr c\302\260slS/rwtan 8b 2crzr c\302\260s\177lrwfb tan,\"

where azr is the vertical stress on the reinforcement element (at depth zr) in the vicinity of the

resisting zone. If the ground surface is horizontal and there is no surface loading, or the sur-
face loading is uniform and covers a large area behind the facing panel, we may assume

07. = Ozr. Thus, equating Equations 8.78 and 8.79 we obtain:

kasxSz
lr >
(8.80)
2wfb tan,\"

The tensile force in the reinforcement

or smaller must be equal
than Tall (Equation 8.75). to

For continuous grid reinforcement, sx and w are unity

and Pt represents the tensile force per
unit length of the walt. For reinforcement of low extensibility, Juran & Schlosser (1978) sug-
gestedan active zone that is separated from the resisting zone by a logarithmic spiral surface
and represents the maximum states of tensile stress in the horizontal reinforcement. The as-
sumed failure surface intersects the ground surface at a distance of 0.3H from the facing panel.
A.ASHTO (1997) suggested that the logarithmic spiral may be idealized by two linear failure

\177
+ e\"
\177z45\302\260 12 Active t Resisting
, \177 zone
\177 \177
Shear resistance
Facing ,, \302\242\177
', Failure plane \177 due to friction

'
t\" t,-\" r\177 !Reinforcement

Figure 8.39. Soil-reinforcement bond due to friction.)))

 Lateral Earth Pressure and Retaining Walls 357

.2 1.7 2.5
k/l%

Reinforcement
Metal\177
strips 1

H1/2 I ,,
Metal bars,
J 6.0m welded wire
a 1
grids
\177ne

H\177/2 i
i
sz

(a) (b)
Figure 8.40. Reinforcement of low extensibilit\302\245: (a) active zone, (b) variation of k / k,\177with depth.

planes (Figure 8.40(a)). The distribution of the appropriate earth pressure coefficientis shown
in Figure 8.40(b). A slightly different arrangement is suggested by BS 8006 (1995). In the
sloping backfill the magnitude of Hj defined in Figure 8.40(a) is:

0.3 H tan \1773

H\177
= H + (8.81)
1-0.3 tan[3

In the presence of surface loading, the arrangement of the failure planes will change and this

can be determined from the relevant code.

8.8.3 Soil-nailed earth retaining watfs

Nailed soils represent a recent development in ground improvement and the technology of
earth retaining structures. They can be classifiedinto two categories of driven and grouted
nails. Driven nails are small diameter rods that are driven into the soil or drilled holes by
means of hammers or (more recently) by compressed air launchers. In the latter category, the
nail is subjected to the tensile stresswhile being fired into the ground thereby eliminating the
possibility of nail buckling. In the former category, the nails are not pretensioned but are
grouted with or without a grout pressure similar to ground anchors. The main advantages of
nailed retaining walls and slopes are their stability against dynamic loading (Felio et al., 1990)
and their use in excavations as welt as retaining walls. Methods of analysis include using an
empirical method that assumes an idealized active zone of locus of maximum tensile stresses
(Bowles, 1996),a kinematic limit analysis (Juran et al., 1988)and the finite element method
(Unterreiner et al., 1997). In the empirical method, the active zone is defined by Figure 8.40(a)
and Equation 8.81 applies. The distribution of the lateral earth pressure is trapezoidal (Figure)))
358 Soil Meclmnics: Basic Concepts and Engineering Applications

8.35(d)) and the maximum ka that occurs at mid-height is given by:

= 0.65 2 \302\260
_
k
a tan (45 0\"/2) for sands with'lH/c\" >- 20 (8.82)

4c\177
ka = I \177
tan2(45\302\260-0'/2) forc', soils
q\302\242 (8.83)
I ,{Htan(45o_\177,/2)

Equation 8.70 is used to estimatethe maximum bond stress developed on the grouted nails.
Both the finite element and kinematic methods have shown that the application of the trape-
zoidal pressure diagrams (used in braced excavations) provides rational estimates of the work-

lug tensile forces in the nails (Schaefer et al., 1997).

Example 8.18

It is required to design a 9 m high vertical face with a horizontal ground surface and using a
geotextile reinforcement strip of width w = 55 mm and Tut\177 = 150 kN. The soil propertiesare:
c' = 0, \302\242\177'
= 35 \177'
and y = 18 kN/m 3.
The tilt is placed in 0.3 m thick layers. Other known data are:
CRF = 0.40, Fc = 1.0,Fd \177-1.2, Fs = 1.3,fb =- 0.8, and Sx -\177s..- = 0.9 m.
Solution:
From Equation 8.75: Tat t
--
T,t t \303\227
CRF / Fc x Fa \303\227
Fs = 150.0x 0.4/1.0 1.2 x
\303\227 1.3 = 38.5 kN.
Using Equation 8.6 we obtain: ka = 0.271. The position of the potential failure plane is:
= 45.0 \302\260
o\177 + \177'/2 = 45.0 \302\260
+ 35.0\302\260/2 = 62.5
\302\260
from the horizontal.
Using Equation 8.80 we have:
t r = 0.271\303\2270.9\303\2270.9! 2\303\2270.055x0.Sxtan35.0
\302\260
= 3.56 m.
The first reinforcement may be ptaced 0.45 m below the ground surface, and thus the last rein-
forcement at the bottom will be at a depth of 8.55 m. The tensiIe load transferred to this rein-
forcement is calculatedfrom Equation 8.78:

Pr =0.271xlS.0\303\2278.55xl.0\303\2270.9x0.9
= 33.8 kN < Tall = 38.5 kN.
Total Iength of reinforcement (first from the top):
B = 8.55/tan62.5\302\260 + 3.56 =8.0 m. One may use a variable length by considering the I inear

variation of the width of the wedge by depth.

8.9 THEOVERALL STABILITY OF RETAINING STRUCTURES

8.9.1 Slopestability analysis

The overall stability of a gravity or a cantilever retaining wall may be evaluated by a slope

stability analysis (Chapter 9) with outside the wall system and

(sIiding) trial circles that are
possible anchorages (Figure 8.41(a)). A similar method is used for a sheet pile and anchor sys-
tem as shown in Figure 8.41 (b) where the (sliding) trial circles pass through the tip of the sheet
pile. The minimum requirements for the factor of safety must comply with the relevant codes.)))
 Lateral Earth Pressure and Retaining Walls 359

(a) (b)

Wa

I I
R R
(e) (d)
Figure 8,41. Evaluation of the overall factor of safety for retaining slructures.

Bowles (1996) suggesteda minimum factor of safety of 1.2 for the mobilization of the shear
strength on the trial circle. For sheet pile wails, failure modes due to the loss of bond between
the rod and grout, and excessive deformation of the sheet pile, must also be considered.

8.9.2 Sliding block atvalysis

Evaluation of the overall stability of anchored sheet pilesusing the method of the sliding block
was first suggested by Kranz (1953) and subsequently modified by Locher (1969) and Little-
john (1970). The method considers the equilibrium of a wedge type block locatedat the back
of the sheet pile that includes the anchor system, as shown in Figures 8.41(c), (block FBCD,
FB being the failure plane of the active zone), and 8.41(d), (block FABC). The lower sliding
surface passes through the zero shear point F in the equivalent sheet pile beam. This sliding
surface terminates at the tip of the vertical anchor or at the mid-point of the grouted anchor

(point C in Figure 8.4 l(d)). A failure mode in which the sliding surface terminates at the top of
the grouted portion (point D in Figure 8.41(d)) also has to be considered. This method, with
some modifications to the geometry of the sliding block, can also be appliedto multiple an-
chors (Hanna, 1982).
The application of the sliding block for overall
stability analysis has been acceptedby the
codes of several European countries 8081, 1989). The sheetpile system
(e.g. BS is assumed to
be a part of the block so that the force and the active thrust are treated as internal
anchor rod
forces that do not affect the stability of the block. Relevant forces are the effective weight of
the block, the soil reaction R on the sliding surface, and the active thrust Pa\177, on the right side
of the anchor which is calculated using a linear distribution of earth pressure. Force equilib-
rium yields both the magnitude and direction of the reaction force R. The overall factor of
safety is defined by the following equation where 0'm is the angle between the reaction force R)))
360 Soil Mechanics: Basic Concepts and Engineering Applications

and a line normal to the sliding surface.

tan
F\302\260verall
= \177\" > 1.5 (8.84)
tan

In the revised version by Ostermayer the passive force acting

(1977), between point F and the
dredge line (if point F is under the dredge line) was included in the force equilibrium. The
overall factor of safety is calculated from Equation 8.84 with a minimum value of 1.2. BS
8081(1989)suggests that the active thrust above point F and anchor rod force be included in
the force equilibrium equation. Note that the horizontal component of Pa (abovepoint F) is

equal and in the opposite direction to the horizontal component of the anchor rod force as the
point F represents the zero shearpoint. In this case the minimum requirement for the factor of
safety defined by Equation 8.84 is 1.3 for non-critical applications and 1.5 for critical applica-
tions. In a stratified soil the sliding surface passes through different types of soil and the prob-
lem is indeterminate unless a simplifying idealization is made. The sliding block is divided
into vertical slices by drawing vertical lines through the points of intersection of the sliding
surface and layers of the soil. From the force equilibrium of each slice, together with the as-

sumption that the angle \177' (for each soil) is fully mobilized on the sliding surface, the corre-

sponding soil reaction R is computed. The factor of safety is defined as the ratio of the com-
puted anchor rod force from the force equilibrium of the block to the anchor rod force
calculated from the statics of the sheet pile.

Example 8.19
Calculate the overall factor of safety for the sheet pile system of Example 8.13. Assume the
length of the anchor rod is 13 m and is supported by a plate anchor of length 2.6 m with its
mid-point 1.3 m below the ground surface.
Solution:
The point of zero shear (F) is located 4.3 m below the water table (see Example 8.13).
Referring to Figure 8.42, the forces applied to the sliding block are calculated as follows:
The lateral active pressure at point F is:
I9.1 + 10.4\303\2274.3\303\2270.286 = 31.9 kPa
Horizontal active thrust above point F:
Pal,
= (6.9 + 19.1)/2x2.6\303\2271.0 + (19.1+ 31.9)/2\303\2274.3xi.0 = 143.4 kN,
(equal to the anchor rod force).
The vertical component of the active thrust is:
= = \302\260
Pay PaI\177
tan8\"
\303\227 I43.4\303\227 tan 17.0 = 43.8 kN.
The horizontal active thrust acting at the back of anchor plate is:
= (6.9 + 19.1)/2\303\227 1.0
2.6\303\227 = 33.8 kN.
Ponl\177

The vertical component is:

= 33.8\303\227tan 17.0 \302\260
= 10.3 kN.
P, nv
Calculate the weight of the sliding block including the vertical load due to the surcharge:
W = 16.5(2.6 x 13.0x1.0)+ 10.4(4.3 x 13.0t 2)x 1.0 + 24.0\303\22713.0x 1.0 = 1160.4 kN.)))
 Lateral Earth Pressure and Retaining Walls 361

24 kPa

6.9 kPa \177, \177 \177r\177 \177 \177 \177 \177

1\177\177
6.9kPa

Bar
2.6 m \177
d
-\177 \" 19.1 kPa

19.\177i
Pay
7.4 m
31.9 kPa

Figure 8.42. Example 8.19.

The sliding block and the corresponding forces are shown in Figure 8.42:
-t
ct = tan (4.3/13.0) = l&3 \302\260.
Horizontal equilibrium:
R sin(O',,,- 18.3\302\260)
-33.8 + 143.4 - 143.4= 0.0.
Vertical equilibrium:
R - 18.3 -10.3
\302\260) + 43.8 - 1160.4-= 0.0, solving for R and \177'm:
cos(d\177n

R = 1127.4 kN, \177'm

= 20.0\302\260.

\302\260
=
From 8.84: = tan 30.0\302\260/tan 20.0 1.59.
Equation Foveral I

8.10 PROBLEMS

8.1 An 8 m high retaining wall retains a soil with two 4 m thick layers that have the follow-
ing properties:
Upper layer: c' = 10kPa, = =
\302\242'18 \302\260,
y 18 _kN/m3;
lower layer: c =0, \177 = 35 , )' = 18 kN/m . For a surface load q = 50 kPa, determine the
active thrust and its distance from the base of the wall.
Answers: 294.7 kN, 3.41m
8.2 A retaining wall of 5 m height retains a sloping backfill with = 20 \302\260,
\1773 The properties of
backfill are:
c' = 0, 4' = 35\302\260,
and 7 = 17 kN/m 3.

Determine the active thrust on the wall and its horizontal and vertical components.
Answers: 68.3 kN, 64.2 kN, 23.4 kN
8.3 Re-work Problem 8.1 assuming that the water table is located 2 m below the ground sur-
face. The saturated unit weight for both layers is 19.5 kN/m 3.)))
362 Soil Mechanics: Basic Concepts and Engineering Applications

Answers: 249.9 kN (note this is not the total horizontal Ioad applied to the wall), 3.62 m
8.4 A 10 m retaining wall retains soil with the following properties:
cu =40 kPa, y --- 17.5 kN/m 3, cw = 17.6 kPa, and \177i'=0.
Determine the magnitude of the active thrust:
(a) when the surface carriesno surcharge,
(b) when a surface surcharge of 50 kPa is applied,
In both cases the ground surface is horizontal.
Answers: 178.3 kN, 475.5 kN

8.5 A retaining waII of height 12 m retains a two-layer soiI the following properties:
having\177
0 m - 5 m below the surface: c\177,\177[2 kPa, @u = = 17 kN/m-.
0,\177
BeIow 5 m, cu = 35 kPa, \302\242Pu
= 0, and
y
= 18 kN/m'.

Calculate the magnitude of the total active thrust and the critical value of ct. For this pur-

pose formuIate Pa in terms of the angle c\177using the force diagram and set:
OPa l Oc\177to zero.
6' = 0 and c\177.
= 10 kPa.

Answers: c\177=40.6 Pa
\302\260, =577.8 kN
8.6 A gravity concrete retaining waII is 6.6 m high and 3.2 m wide. if the thickness of the
soil at the front of the wall is 2 m, determine the maximum and minimum base pressures
assuming no base friction or adhesion. ThesoiI l\177as the following
prol\177erties:
c\" = 0,
0' = 35% 13 (for soil) = 1.8 Mg/m3, p (for concrete)= 2.4Mg/m \177.

Answers: 61.8 kPa (at the heel), 249.0 kPa (at the toe)

8.7 For the gravity concrete

retaining wall of Problem 8.6, determine the maximum and
minimum base pressuresassuming a base friction angIe of 6' = 15\302\260,
and a base adhesion
of 10kPa. A surcharge load of 20 k_Pa is applied vertically to the ground surface on tt\177e
backfiIl. The soiI has the following properties:
c' = 10 kPa, 0' = 25 \302\260,
p (for s_oii)
= 1.8
Mg/m 3.
p (for concrete)= 2.4Mg/m \177.

Answers: 38.8 kPa (at the heel), 271.9 kPa (at the toe)

8.8 A gravity concrete retaining wail with the soiI of Example

a vertical 8.4.
back supports
The width of the wall is 3.5 m
base and 2 m at the at the and has a 2 m vertical front
to\177
c' =20 kPa, \177'= 25 and y = 18 kN/m . The rest of the front
\302\260
supported by soil with face is

sloping upwards, terminating at the ground surface. Determine:

(a) the factor of safety against sliding assuming that the cohesion between the base of the
wall and the soil is 20 kPa, and the mobilized friction angle on this interface is 25 \302\260,
(b) the factor of safety against overturning,
(c) t[\177edistribution contact pressure under
of the the base of the wall.

Take the unit weight of the concrete as 24 kN/m 3. Assume the back and front faces of
the walI are smooth (c.. = 0, \177'= 0).

Answers: i\17765, t.61, -91.1 kPa (at the heel), 338.0 kPa (at the toe))))
 Lateral Earth Pressure and Retaining Walls 363

8.9 A cantilever sheet ile supports a 6 m hig\177h backfill with the following properties:
p,
0m-2m:c'=0,0 =30 ,\302\245=16.5kN/m a, \302\260

2 m - 4 m: c' = 0, 0' = 35\302\260,

Y = 17 kN/m 3,
4 m - 6 m: c' = 15 kPa, lp' = 20 \302\260,
\302\24517 kN/m 3. D
= = 3.5m and the soil under the dredge
line (at both sides) is the same soil under
as the the 4 m depth.
Determine the factor of safety (in terms of c' and
kp) assuming a simplified net pressure
diagram.
Note that the case of the cantilever sheet pile in c', \177' soil is not formulated in this chap-
ter.
Answer: 1.22
8.10 A cantilever sheet pile supports a 9m high backfill with the following properties:
0m- 3 m: c' =0,0'--35\302\260,y
= 18 kN/m 3,
3 m - 6 m: c' = 15 kPa, \177'= 20 \302\260. =
Ys,\177t 20.3 gk.N/m 3,
6 m - 9 m: c' = 0, 0\" = 35\302\260,
Ysaz
= 21.1 kN/m 3.
Thereis a vertical surface load of 20 k.Pa applied at the bacH\177ill. The water table is 3 m
below the ground surface of the backfill and has the same level at the front of the sheet
pile.The soil under the dredge line (at both sides) is purely cohesive soil with:
cu= 100 kPa, and 0, = 0.
(a) Determine the embedment depth D assuming that the full passive resistance is mobi-
lized,
(b) add anchor rod at a depth of 1.5 m and with the same embedment depth
a horizontal
of part (a) calculate the factor of safety in terms of mobilized cohesion under the dredge

line and the anchor rod force,

(c) determine the location and magnitude of the maximum bending moment,
(d) if the anchor rod is supported by a concrete block anchor with thickness of 0.5 m,
width of 2 m (parallel to the sheet pile), and height of 1.5 m, calculate the distance be-
tween anchor rods along the sheet pile (anchor rod is anchored at the centre point of the
concrete block).
Assume no cohesion and friction resistance along the surfaces of the anchor. Include the
surface load in both the active and passive thrusts and assume a factor of safety of 1.5 for
the mobilized kp at the front of the anchor.
Answers: 4.1 m, 2.5, 119.4kN/m, x = 0.42 m (below the second layer),
BM = 289.7 kN.m/m, 2.6 in
8.11 An earth retaining wall 12 m higl\177 is reinforced with metal strips of width 1130 ram, the

first row of which is at a depth of 0.5 m. The strips are spaced at sx= 1 m and s:
= 1 m,
and the allowable tensile strength of the metal strip is 140 MPa.
The properties of backfill soil are:
c\" =
0, 0' = and
35\302\260, y = 18 kN/m 3,
The friction angle mobilized in the soil-strip contact area is \177/,
= 23 \302\260.
Calculate the
length of the reinforcements at depths 0.5 m, 5.5 mand 11.5 m.

Answers: 8.9 m, 7.6 m, 4.2 m)))

364 Soil Mechanics: Basic Concepts and Engineering Applications

8.11 REFERENCES

AASHTO 1997.Standard specifications for highway bridges. Section5.8.Washington, D.C.

AIpan, 1. 1967. The empirical evaluation of the coefficient ko and ko\177oce. Soils and foundations 7(1): 31-
40,Tokyo.
ASTM D-5262. 1997. Standard test method for evaluating the unconfined tension creep behavior for
geosynthetics. Philadelphia: ASTM.
Benoit, J. & Lu\177enegger, A.J. 1993. Determining lateral stress in soft clays. In G.T. Houlsby & A.N.
Schofield (eds). soil mechanics. London:
Predictive Thomas Tel ford.
Bishop, A.W. 1958. Test requirements for measuring the coefficient of earth pressure at rest. Proc. Brus-
sels conf on earth pressure problems.
Bowles, J.E. 1996. Foundation analysis and design. New York: McGraw-Hill.
BS 8006. 1995.Code of practice for strengthenedCeinforced soils and other fills. London: British Stan-
dard Institution.
BS 808I. 1989. Ground anchorages. London: British Standard Institution.
BurIand, J.B., Ports, D.M. & Walsh, N.M. 1981. The overall stability of free and propped embedded can-
ti]ever retaining wails. Ground engineering 14(5): 28-38.
Cheney, R.S. 1988. Permanent ground anchors. Report FHWA/DP-68/IR U.S: 136. Washington, D.C.:

Department of Transportation Federal Highway Administration.

Das, B.M. 1990. Earth anchors. Amsterdmn: Elsevier.
Das, B.M., Tarquin, A.J. & Moreno, R. 1985. Model
tests for pullout resistance of vertical anchors in
clay. Civil eng. for practising and design engineers 4(2): 191-209. New York: Program Press.
Duncan, M., Clough, G.W. & Ebeling, R.M. 1990,Behavior and design of gravity earth retaining struc-
tures. In P.C. Lambe & L.A. Hansen (eds), Design and performance of earth retaining structures:
251-277. New York: ASCE.
Duncan, J.M., & Seed, R.B. 1986. Compaction-induced earth pressure under /\177-conditions. Journal
SMFE. ASCE 112(]): 1-22.
ncl
Exxon. 1992. Designing for soil reinforcement. 2 edition, UK: Exxon Chemical Geopolymers.
Felio, G.Y.,Vucetic, M., Hudson, M., Barar, O., & Chapman, R. 1990. Performance of soil nailed walIs
r\177f
during the October 17, 1989Loma Prieta earthquake. Proc. 43 Canadian geotechnicat conf.: 165-
173. Quebec.
Finale, K. Peck, R.B. 1973. Braced cuts in sand and clay. Norwegian geotechnicaf institute, Publication
96.
Fleming, W.G.K., Weltman, A.J., Randolph, M.F.& Elson, W.K. I992. Piling engineering. Surrey Uni-

versity Press: Halsted Press.

st
Hanna, T.H. 1982. Foundations in tension-ground anchors. 1 edition: 269-274. USA: McGraw -Hill.
th
HueckeJ, S. 1957.Model tests on anchoring capacity of vertical and inclined plates. Proc. 4 intern.

conf. SMFE 2: 203-206.London.

n'\177
JLky, J. 1948. Earth pressure in soils. Proc. 2
intern, conf SMFE 1:103-107. Rotterdam.
Juran, I. & Schlosser, F. I978. Theoretical analysis of faiIure in reinforced earth structures. Proc. syrup.
on earth reinforcement, ASCE convention, Pittsburg: 528-555. New York: ASCE.
Juran, I., Baudrand, G., Farrag, K., & Elias, V. 1988. Kinematical limit analysis for design of nailed
structures. Journal GED, ASCE, 116(I):54-72.
fcl
Koerner, R.M. 1994. Designing with geosynthetics. 3 edition, Englewood, New Jersey: Prentice Hall.
Kranz, E. 1953. Ober d& verankerung yon spundwfinden. Berlin: Verlag Ernst & Sohn.)))
 Lateral Earth Pressure and Retaining Walls 365

Kulhawy, F.H., Jackson, C.S. & Mayne, P.W. 1989. First order estimation of ko in sands and clays.
Foundation engineering: current principles and practices, Journal GED, ASCE, 1 : 121 - 134.

Littlejohn, G.S, 1970.Soil anchors. Proc. conf ground engineering: 33-44.London: Institution of Civil
Engineers:
Littlejohn, G.S. & Bruce, D.A. 1977.Rock anchors-state of the arts: 50. UK: Foundation Publications
Ltd..
Locher, H.G. 1969.Anchored retainh\177g walls and cut.offwalls. Berne: Losinger & Co..

Mayne, P.W. & Kulhawy, F.H. 1982. ko-OCR relationships in soil. Journal GED, ASCE, 108(GT6):851-
872.

Meyerhof, G.G. 1973. Up/ift resistance of inclined anchors and piles. Proc. intern, conf SMFE2.I: 167-
172. Moscow.
Munfakh, G.A. 1990.Innovative earth retaining structures: Selection, design, & performance, In P.C.
Lambe & L.A. Hansen (eds), Design and performance of earth retaining structures: 85-118. New
York: ASCE.
Munfakh, G.A., Abramson, L.W., Barksdale, R.D. & Juran, I. 1987, Soil improvement-a \177enyear \177pdate.
In J.P. Welsh (ed), ASCE geotechnical special publication (12): 59. New York: ASCE.
O'Rourke, T.D. 1987.Lateral stability of compressible walls. Geotechnique 37(1): 145-149.
O'Rourke, T.D. & Jones, C.J.F.P. 1990.Overview of earlh retention systems: 1970-1990. In P.C, Lambe

& L.A. Hansen (eds), Design and performance of earth retaining structures: 22-51. New York:
ASCE.
Ostermayer, H. 1977,Practiceon detail
design application of anchorages-A
the review of diaphragm
walls: 55-6t. London: Engineers.
Institution of Civil
Ovesen,N.K. & Stromann, H. 1972. Design methods for vertical anchor slabs in sand. Proc. speciality
conf on performance of earth and earth supported structures 2(1 ): 1481-1500, New York: ASCE.
Palossy, L., Scharle, & Szalatkay, 1. 1993. Earth walls. New York: Ellis Horwood.
\177
Peck, R.B. 1969. Deepexcavation and tunnelling in soft ground. Proc. 7 intern, conf SMFE 225-290.
Mexico.
Peck,R,B. 1990.Fifty years of lateral earth In P.C. Lambe &
support. L.A. Hansen (eds), Design and
performance of earth retaining structures. New York: ASCE.
Post-tensioning Institute (PT1). 1996. Recommendationsfor prestressedrock and soil anchors. Phoenix,
Arizona,

Pot,s, D.M, & Fourie, A.B. 1984.The behaviour of a propped retaining wall: results of a numerical ex-
periment. Geotechnique 383-404.
34(3):
Ports, D.M. & Fourie, A,B. 1985. The effect of wall stiffness on the behaviour of a propped retaining
wall. Geotechnique 35(3): 347-352.
Rowe, P.W. 1952. Anchored sheet pile walls. Proc. institution of civil engh\177eers, 1(1): 27-70.
Rowe, P.W. 1957.Sheet pile wails in clay. Proc. institution of civil engineers, 1(7): 629-654.
Schaefer,V.R., Abramson, L.W,, Drumheller, J.C., Hussin, J.D. & Sharp K.ID. 1997. Ground improve-
ment, ground reinforcement, ground treatmem-developments 1987-I997. ASCE geotechnical special
publication (69): 178-200. New York: ASCE
r\177t
Scott, C,R. 1980,An introduction to soil mechanics and foundations. 3 edition. London: Applied Sci-
encePublishers.
Task force 27 (Federal Highway Administration). 1990. ln-sim soil improvement \177echniques: design
guidelines for use of extensible reinforcements (geosynthetics) for mechanically stabilized earth walls
in permanent applications. Joint committeeof AASHTO, AGC, ARTBA. Washington, D.C.
Teng, W.C, 1962, Fou\177wlation design. Englewood Cliffs, New Jersey: Prentice-Hall.)))
366 Soil Mechanics: Basic Concepts and Engineering Applications

t\177
Ter-zaghi,K. 1966. Theoretical soil mechanics. 14 edition. New York: John Wiley & Sons.
nd
Terzaghi, K. & Peck,R. B. 1967. Soil mechanics in enghzeering practice. 2 edition, New York: John
Wiley & Sons.
r\177
Terzaghi, K., Peck,R. B., & Mesri G. 1996. Soil mechanics in engineering practice. 3 edition, New
York: John Wiley & Sorts.
n\177t
Tschebotarioff, G.P. 1973, Foundations, retaining and earth structures, 2 edition. New York:
McGraw-Hill,
Unterreiner, P., Benhamida, B. & Schlosser, F. 1997. Finite element modelling of the construction of a
full-scale experimental soil-nailed wall, French national research project CLOUTERRE. Jo\177rnat of
ground improvement 1 (i): I-8.
Williams, B.P. & Waite, D. 1993. The design and construction ofsheetLpited cofferdams. Special publi-
cation 95. London: Construction Industry Research and Information Association,
Williams, G.W., Duncan, J.M. & Sehn, A.L. 1987.Simplified chart solution of compaction- ittduced
earth pressures ott rigid structures. Geotechnicat engineeri\177g report, Blackburg, VA.: Virginia Poly-
technic Institute and State University,
Wroth, C.P. & Houisby, G.T. 1985. Soil mechanics-property characterization and analysis. Proc. intern.

conf SMFE 1: 1-56.

Xanthakos, P,P. 1991.Ground anchors and anchored structures. New York: John Willey & Sons.
Xanthakos, P.P., Abramson, L,W. & Bruce, D.A. 1994.Handbook on ground control and improvement.
New York: Iohn Willey & Sons.)))
CHAPTER 9

Stability of Earth Slopes

9.1 INTRODUCTION

This chapter examines the stability of earth in two-dimensional space using

slopes a limit

equilibrium method. Figures 9.1(a) and 9.1(b) a rotational failure

illustrate where a mass of
soil rotates on a circular or non-circular failure surface at its limit state. In the presence of a
hard layer, a combined translationaI and rotational faiIure could occur (Figure 9. l(c)).An infi-
nite slope is normally associated with a translational failure parallel to the ground surface
(Figure 9. l(d)). The failure of a reinforced slope, however, is rather different and is idealized
either by a combination of wedges (Department of Transport UK, 1994) or by a circular fail-
ure surface (Koerner & Robins, 1986; Koerner,1994).
In a rotational circular failure, the mass of soil is treated as a rigid body sliding on the fail-

ure surface (Fellenius, 1927;Bishop, 1955).On this surface the Mohr-Coulomb failure crite-

rion applies and the shear strength parameters used correspond to the peak strength obtained
by a total or effective stress analysis, For a frictional soil, the mass is divided into vertical
slices to facilitate the application of the force and moment equilibrium requirements. In Fel-
lenius' method, only moment equilibrium is satisfied, but Bishop's simplified method satisfies

'
(a) (b)

(c) (d)

Figure 9. I. Idealized types of ear[h slope failure in two dimension,

367)))
CHAPTER 9

Stability of Earth Slopes

9.1 INTRODUCTION

This chapter examines the stability of earth slopes space using a Iimit
in two-dimensionaI

equilibrium method. Figures 9.1(a) and a rotational failure

9.1(b) illustrate where a mass of
soil rotates on a circular or non-circular failure surface at its limit state. In the presence of a
bard layer, a combined transIational and rotational failure could occur (Figure 9. l(c)).An infi-
nite slope is normally associated with a translational failure parallel to the ground surface
(Figure 9, l(d)). The failure of a reintbrced slope, however, is rather different and is ideaIized
either by a combination of wedges (Department of Transport UK, 1994) or by a circular fail-
ure surface (Koerner& Robins, 1986; Koerner, 1994).
In a rotational circular failure, the mass of soil is treated
as a rigid body sliding on the fail-
ure surface (Fellenius, 1927;Bishop,1955).On this surface the Mohr-Coulomb failure crite-
rion applies and the shear strength parameters used correspond to the peak strength obtained
by a total or effective stress anaIysis. For a frictional soil, the mass is divided into vertical
slices to facilitate the application of the force and moment equilibrium requirements. In Fel-
lenius' method, only moment equilibrium is satisfied, but Bishop's simplified method satisfies

crack
i
(a) (b)

(c) (\177)

Figure 9.1. Idealized types of earth slope failure in two dimension.

367)))
368 Soil Mechanics: Basic Concepts and Engineering Applications

both moment equilibrium and vertical force equilibrium. In non-circular failure surfaces, all of
the methods introduced by Morgenstern & Price (1965), Spencer(1967)and Janbu (1968 &
1973) satisfy both force and moment equilibrium, but differ slightly in the assumptions made
for the side forces acting on a slice. Circularfailure surface analyses yield results that are suf-
ficiently accurate for many practical purposes.
The factor of safety is defined as the ratio of the shear strength to the mobilized shear stress
on the sliding surface required for equilibrium, and is assumed to be constant along the sur-
face. Whilst the factor of safety actually varies along the sliding surface (Tavenas al., 1980);
et

the average value obtained from a traditional circular analysis is a reliable indication of overall

slope stability.
The choiceof drained or undrained conditions depends upon the magnitude of the time fac-
tor Tv defined by Equation 6.20. For the construction time t substituted into this equation, if
the value of Tv exceeds 3, it is reasonable to treat the material as drained. If the value of Tv is
less than 0.01, the material is treated as undrained (Duncan, 1996). For the values of Tv be-
tween the above limits both drained and undrained analyses must be considered. In the ab-
senceof consolidation data, consideration should be given to the magnitude of the coefficient
of permeabd\177ty k. So\177Is wtth values ofk than 10 m/s can be assumed to be drained,
\177reater
whilst those with values ofk less than 10- m/s can be considered to be undrained. In both cut
and fill slopes, it is customary to study their short-term stability in undrained conditions with

shear strength parameters cu, #,; their long-term stability, however, should be examined using
the effective shear strength parameters c', \302\242/(Atkinson, 1993).
For homogenous soils, general solutions for the critical values (at failure) of the dimen-
sionless parameters yH / c\" or c\" / yH have been obtained for specific values of \177)'and the slope
angle \1773(Taylor, 1948; Bishop & Morgenstern, 1960;3ranbu, 1968 & 1973; Cousins, 1978;
Duncan et al., 1987). For slopes reinforcedby geosynthetics, general solutions and charts pro-
vided by Jewell et al. (1985),Jewell (1991 and 1996) can be used. In engineering practice
however, an individual solution is usually required to acconmaodate geometric irregularities,
variations in the shear strength parameters and different layers of soils. In a slope stability
analysis the objective is to locate the circle (or any type of the sliding surface) that yields the
minimum factor of safety. This usually requires \177hat an iterative procedure be employed.

9.2 STABILITY OF SLOPES IN c,, Ou

= 0 SOIL - CIRCULAR FAILURE SURFACE

9.2.1 Factor of safety in c,, 9u = 0 soil using circular failure surface

A plane strain slope of height
angle H
!3 is shown in Figure
and 9.2(a). The trial circular fail-
ure surface is defined C, radius R and central angle 0. Shear stressesalong
by its centre the
trial surface are due only to undrained cohesion and are mobilized to cu / F (to anaintain the

equilibrium of the sliding block), where F is the factor of safety. The weight of the sliding
block W acts at a distance d from the centre of the circle. Taking moments of the forces about
the centre of the circular arc, and noting that the normal stresses on the arc pass through the
centre, then:)))
 Stability of Earth Slopes 369

I\177 XC \1771\177

\177\" '
/ \\ _\302\242
\177 \177 ,,.

l Yc--,- / B f\177

(a)
= 0
l:igure 9.2. ,Slope failure in c,,, \302\242, soil.

20
cu R
=
cu L a R
F = (9.1)
Wd Wd

where La is the length of the circular arc. If a surcharge loading q is applied to the upper-
ground surface and external water exists at the front of the slope (Figure 9.2(a)):

F= cuR20
(9.2)
Wd + qLs - Pwe
t

If the soil is composed of two different layers (Figure 9.2(b)), F is obtained from:
R (cul0 t +cu202)
F = (9.3)
Wd
where cu] and cu2 are the undrained cohesions of soiI 1 and soil 2 respectively and O'l and 0'2
are the corresponding central angles. The stability of an earth slope in undrained conditions
can be expressedin terms of a dimensionless parameter N called the stability number:

N =
yH / cu (9.4)

It can be shown that for a specified value of J3, the magnitude of N at failure has a constant
value (Nf) and, as a consequence,the factor of safety (Equations 9.1) may be presented by:
F =
Nf I N d (9.5)

where Nd is number the corresponding

stability to the design values of y, H and cu. The cen-
troid of the
sliding mass is obtained using a mathematical procedure based on the geometry or
the sub-division of the sliding mass iqto narrow vertical slices.)))
370 Soil Mechanics: Basic Concepts and Engineering Applications

Example 9.1

Find the factor of safety of a i vertical to 1.5 horizontal slope that is 6 m high. The centreof
the trial circle is located 2.5 m to the right of and 9.15 m above the toe of the slope, cu = 25
kPa, and \177= 18 kN/m 3,
Solution:
Geometrical data are:0 = 85.9
2
area
\302\260, of the sliding mass = 29.87m and d = 3.85 m.

R = \177/9.152 + 2.52 = 9.48 m, W = 29.87 \303\227 18.0

1.0\303\227 = 537.7 kN.
F = cz, R 20 ! Wd = I25.0\303\227
9.482 \303\227
(85.9
\302\260
/ 180.0\302\260)n]/(537.7 \303\227
3.85)
= 1.63.

9.2.2 The effect of tension cracks

Tension cracks may develop from the upper ground surface to a depth Zo that can be estimated

using Equation 8,12. The effect of a tension crack can be taken into account by assuming that
the trial failure surface terminates at the depth zo, thereby reducing the weight W and central

angle 0. Any external water pressure in the crack creates a horizontal force that must be in-
cluded in equilibrium considerations.

Example 9.2
Re-work Example9.1by taking into account tension cracks.
Solution:
UseEquation 8.12: zo =
2c u /\302\245
= 2 \303\227
25.0/1. 8.0 = 2.78 m,
From the geometry of the circle and slope (calculations omitted): 0 = 66.6\302\260,
area of the sliding
2
mass = 27.46m and d = 3,48 m.
18.0 = 494.3 kN. Calculate the
W \177-27.46\303\2271.0 \303\227 horizontal force (Pw) due to the water pres-
sure in the tension crack and its vertical distance (Yt) from the centre of the circle:
= 9.81.\303\2272.782/2
P,\177.
= 37.9 kN, Yt
=
Yc
- H + 2z
o/3 = 9.15-6.0+ 2\303\2272,78/3 = 5.0 m.
Parameters zo and yc e defi\177led
a\177 in Figure 9.2(a). Taking moments about the centre:

F = cu R20/(Wd + PwY\177 )
= ( 25.0\303\2279.482 494.3 \303\2273.48+ 37.9\303\2275.0)
\303\22766-6\302\260\303\2277\177/180\302\260)/(
= 1.37.

9.2.3 Location of the critical circle

Cousins (1978) developed a seriesof charts for homogeneous soils using extensive computer
analyses to investigate the effects of pore pressure and a hard stratum (located horizontally
under the slope) on the magnitude of the stability number N. Figure 9.3 is reconstructed for a
c,\177,Ou
= 0 soil where the hard layer is deep and has no effect on the mode of failure. The coor-
dinate of the centre of the critical circle through the toe (xc, Yc) in the xAy pIane (with its ori-

gin A at the toe) is normalized in terms of height and slope angle. Note that the most critical
circIemay or may not be a toe circle.Selecting a trial circIe using Figure 9.3 or other availabIe
charts can reduce the number of iterations in the search for critical circle.)))
 Stability of Earth Slopes 371

/ H
1.4 \177an\177
yc

tanl3 xc/1-1 1.2

& 1
0.8
tanl\177 y\302\242/H
0.6 j /

02 tan\177

0 5 10 15 20 25 30 35 40 45

\177(degrees)

Fibre 9.3. The position of the cddca] \177 circle in cu, \177,
= 0 soil (Cousins, ]978).

Example 9.3
A cut i0 m deep is to be made in a stratutn of cohesive \177il for which c,, = 45 \177a. The slope
\177
angle =
\177 40 and the soil h\177 a unit weight of 17 \177/m 3. Using Cousins's cha\177 of Figure 9.3,
find the factor of safety for the c\177gcal toe circle.

Solution:
From Figure 9.3: tan\177xc / H = 0.50, tan\177yc / H = 1.2 i.
xC = 0.5\177 10.0/ \177an40.0
\177
= 6 =
m, Yc ['2[xl0.0/tan40-0\177 \17714.4 m.

The central angle, area of the sliding block and the position of its centroid are f\177und to be:
O = 96.2% S= 128.5m 2, d= 5.24 m.
R = \1776.02 + 14.42 = t5.6 m, W = 128.5x 1.0x 17.0 = 2184.5\177. From Equation 9.1:
F = (45.0 x 15.62 x96.2x \177/180
\177
\302\260)/(2184.5 x5.24)
= 1.60,

9.2.4 Taylor's stability charts

In undrained conditions, a horizontal hard stratum located ndH below the upper ground surface
affects the critical stability number Nf.
The stability number increases as nd decreases. For 13 >
53 the
\302\260, critical circle is a toe circleand tbe hard stratum has no effect on the stability number.
The slope stability chart developed by Taylor (1948) is shown in Figure 9.4, where the dashed
curves represent the undrained conditions.

Example 9.4

Using Taylor's stability chart of Figure 9.4, determine the factor of safety for the slope of Ex-
ample 9.3.
Solution:
For 13 = 40 the
\302\260, stability number from the chart is 5.52.)))
372 Soil Mechanics.. Basic Concepts and Engineering Applications

12

11

10

9
t \177 t 4.0
8

7
6
.. \17768o
=
n = \177
....
5.52
5 \177 d

\177
4
3.8
3
90 80 70 60 50 40 30 20 10 0
[3 (degrees)

Figure 9.4. Relationship between stability number and slope angle (Taylor, 1948).

N
d yH lc, = 17.0x10.0/45.0=
= 3.78, F = 5.52/378
thus = 1.46. This is tess than the 1.60
obtained in Example 9.3, indicating that the selected circular surfacewas not the critical one.

9.3 STABILITYOF SLOPES

IN c', \177\"SOIL
- THE METHOD OF SLICES

9.3.1 Forcesacting on a slice and moment equilibrium

The soil mass above a trial circle is divided into a series of vertical slices of width b as shown
in Figure 9.5(a). For each slice, its base is assumed to be a straight line defined by its angle of
inclination with
o\177 the horizontal whilst its height h is measured along the centreIine of the
slice. The forcesacting on a sIice shown in Figure 9.5(b) are:
w = total of slice = h x b x where is the
weight the y y unit weight,
N = total normal force on the base = N\" +U, where N'is the effective normal force and U = ul is
the force due to the pore pressure at the midpoint of the base of length l,
T= mobilized shear force on the base = \"cml, where '\177mis the shear stress required for equilib-
rium and is equal to the shear strength divided by the factor of safety: % = \"of/F,
XI, X2 = shear forces on sides of slice, El. E2 = normal forces on sides of slice. Due to the in-
ternal nature of the side forces the sum of their moments about the centre is zero. Thus for
moment equilibrium about C:
i=, i=n
(z ft)i
= = R - i= i, 2,...n, where n is the total number
\177.,TiR R\177.,(Zml)i \177_\177(wsincQiR,
i=1 i=1 i=l f i=1)))
 Stability of Earth Slopes 373

Yc

Figure 9.5. The method of slicesfor c', O\" soil.

of slices. Replacing \"of by c'+\177n tan#' (Mohr-Coulomb failure criterion) we obtain:

i=tt

2[(c' +\177;
tan
2(c'I
+ N\"
tan,')/
q\177')/]i
i=1
F- , F = (9.6)
\177(wsin \177)i \177 (wsin
i=1 i=1

The term c'l may be replaced by c'b !coscc For uniform c', the algebraic summation of c'f is
replaced by c'La, where La is the length of the circular arc. The value of N' must be determined
from the force equilibrium equations. As the problem is statically indeterminate, some simpli-
fying assumptions have to be made. Two common methods that apply different simplifying

assumptions are Fellenius' method and Bishop's simplified method.

9.3.2 FelIenius' tnethod

This method assumes that for each slice Xl =X2 and El = E2. Resolving the system of forces
normal to the base of each slice: N = wcos\177x, N\" + ul = wcoso\177, and N\" = wcosc\177 - ul. Substitut-
ing N' into Equation 9.6 and assuming ru = ub/w gives:

E[c'l + (wcos a- u/) tan,']/ Y\177lcl + w(cosc\177- r. seccO tan

O']i
F=i=I (9.7)
(wsin
\177'\177 \177x)i y\177(wsin o:) i
i i i=1
The parameter ru is dimensionless because the term ub = Yw
x
h\177.
xbx I represents the weight)))
374 Soil Mechanics: Basic Concepts and Engineering Applications

of water with a volume of hwx b x 1.0,where hw is the height of the water above the midpoint
of the base. Furthermore, r,, can be simplified as follows:
ru = ub / w = ub / btry = tt /
yh = y whw / yh (9.8)
In the of steady seepage, the height of water above the midpoint
case of the base is obtained by
constructing the flow net. Alternatively, hydrostatic conditions may be assumed with an aver-
age rf\177value for the slope. By doing so it is assumed that the height of the water above the base
of each slice is a constant fraction of the average height of each slice. If the height of the water
and the average height of the slice are equal, the maximum value of r\177,becomes / y which is
\"y\177.

approximately 0.5. Note that the effective normal force N' acting on the base is equal to
N' = wcos0t- ul or w(cosot - r,\177sec\177t). If the term cos0t - rusec0t is negative, N' is set to zero
because effective stress cannot be less than zero. The whole procedure must be repeated for
number of trial circles until the minimum factor of safety corresponding to the critical circle is
determined. The accuracy of the predictions depends on the number of slices, position of the

critical circle (shallow or deep) and the magnitude of r,\177.For high values of r,, and circles with
a relatively long radius, a conservative value for the factor of safety is obtained that could be
20% or more on the low side.

Example 9.5

Using Fellenius' method of slices, determine the factor of safety for the slope of Example 9.1
for r,, = 0 and 0.4. Take the number of slices as g, each havi\177g 1.5 metre width (check the
width of the last slice). Soil properties are: c' -- 10 kPa, = 29 \302\260,
\177)\" and y = 18 k-N/m 3.

Solution:

From the calculations in Example 9.1 and the geometry of the circle, it can be shown that
R = 9,48 m and xo = 11.44 m (point D is defined in Figure 9.5(a)). The width of slice 8 is:
-
11.44 7 x 1.5 = 0.94 m. Equation of the trial circle in xAy coordinate system is:
2 2 =
(x- 2.50) + (y-9.15) 9.482. Differentiating this equation with respect to x:
2(x - 2.50)
+ 2(y -
9.15)(dy /d_v) = 0, dy / dx = tan ot = -(x
- 2.50)/(y - 9.15).
For each slice, average values of y, 0t, and h are tabulated below along with the other terms
neededf-or the conventional Fellenius' method. Sample calculation for slice6:

ru=O ru = 0.4
ya\177. h 0\177 w wsino:
Slice w(coso: - rusece\177)
(m) (m) (deg.) (kN) wcoscz
(kN)
(kN) (kN)

I -0.14 0.64 -10.6 17.28 16,98 9.95 -3.18

2 -0.30 1.80 -1.5 48.60 48.58 29.14 -1.27
3 -0.22 2.72 7.6 73.44 72.79 43.16 9.71
4 0.10 3.40 16.9 91.80 87.83 49.46 26.68
5 0+71 3.79 26.6 102.33 91.50 45.72 45.82
6 1.66 3.84 37.3 103.68 82.47 30.34 62.83
7 3.15 2.85 49.9 76.95 49.56 1.78 58+86
8 5.03 0.97 63.2 t 6.41 7.40 0.00 14,65
Total: 457.11 209.55 21&10)))
 Stability of Earth Slopes 375

x6 (x coordinateat the mid point of the arc at slice 6) = 5\303\227

1.5 + 1.5/2 = 8.25 m, substituting
into the equation of the circle we find: Y6 = 1.60 m, thus:
c\177
6
= tan-l[-(8.25 -2.5)/(1.60 -9.15)] = 37.3\302\260.
Find the y value at the middle of the base
(not arc) y6,\177, by taking the average values ofy at the two ends of the base, thus:

h = xi tan 13- Yiav

= -
xi / 1.5 Yiav for x <_ x B
= 9.0 m, and h = 6.0 - Yiav
for x > x B = 9.0 m,
where point B is defined in Figure 9.5(a).
For x6a (x coordinateat the left end of the base) = 5 x 1.5 = 7.5 m, Y6L = 1.09 m.
For x6R (x coordinate at the right end of the base) = 6 x = 9.0
m, Y6R = 2.24 m. 1.5
=
(1.09 + 2.24) / 2 = 1.66m, h = 8.25 / 1.5 - 1.66= 3.84m.
Y6uv 6
w = 1.5\303\2273.84x 1.0xlg.0 = 103.68 kN.
6
= = 10.0(85.9\302\260xrU180.0\302\260)9.48
= 142.13.
\177c't c'\177l
Using Equation 9.7 for r, = 0:F = (142.13 + 457.1 Ix tan 29.0 \302\260)/214.10 = 1.85.

For r, = 0.4: F = (142.13 + 209.55x tan 29.0\302\260)/214.10 = 1.21.

9.3.3 Bishop's simplified method

This method slice XI =X2 but E] # E2. Theseassumptions
assumes that for each are considered
to make this accurate than Fellenius'
method more method. An increase of 5% to 20% in the
factor of safety over Fellenius' method is usually realised. Referring to Figure 9.5(b) and writ-
ing force equilibrium in vertical (in order to eliminate El and E2), an equation tbr N' can be
found and substituted into Equation 9.6. Putting t = b / cosa and ub = wru, we obtain:

N\" =
(w- ul - c\177fsin a/F)/(cosa + sin c\177tan \177'/F), which in turn gives:

l
F=,=\177 \177Ic'b+w(1-ru)tan\177' 1
(9.9)
i=]L m\177 -\177i
\177 (wsin(\177)i
i=1
where mo\177is defined by:
= coso\177+ sin\177tand\177'/F (9.10)
mc\177

Equation 9.9 is non-linear in F and is solved by fixed-point iteration. An initial value for F is
guessed (slightly greater than the F obtained by Fellenius' method) and substituted into the

appropriate terms of Equation 9.9 value for F. This procedureis repeated

to compute a new
until the difference between the assumed F becomes negligible. Convergenceis
and computed
normally rapid and only 2 or 3 iterations are required. The procedure is repeated for number of
trial circles to locate the critical failure surface with the lowest factor of safety. Due to the
computational nature of this method, it can be readily programmed on a computer and ex-
tended to deal with layered soils, irregular slope geometry (Bromhead, 1992) and variable ru
(King, 1989).

Example 9.6
Re-work Example 9.5 for r,= 0.4 using Bishop's simplified method.)))
376 Soil Mechanics: Basic Concepts aszd Engineering Applications

Solution:
A summary of the computations is tabulated below where 3 iterations have been carriedout.
The initial value for F was taken as 1.2. For subsequent iterations, the initial value of F is that
computed from the previous iteration. The final factor of safety for the selected trial circle is
1.34.

(a) (b) (c) F= 1.200 F= 1.319 F= 1.339

Slice c' x b w x (I-ru) x (a) + (b)
tan\177' (kN)
-- (d) (c)/(d) -- (d)
(c)/(d)
-- (d)
(c)/(d)
(kN) (kN) ma rna tna

l 15.0 5.75 20.75 0.898 23.11 0.906 22.90 0.907 22.88

2 15.0 16.16 31.16 0.987 31.57 0.989 31.51 0.989 31.51
3 15.0 24.42 39.42 1.052 37.47 1.047 37.65 1.046 37.69
4 15.0 30.53 45.53 1.091 41.73 1.079 42.20 1.077 42.27
5 15.0 34.03 49.03 1.101 44.53 1.082 45+31 1.080 45.40
6 15.0 34.48 49.48 1.075 46.03 1.050 47.12 1.046 47.30
7 15.0 25.59 40.59 0.997 40.71 0.966 42.02 0.961 42.24
8 9.40 5.46 14.86 0.863 17.22 0.826 17.99 0.820 18.12
Total = 282.37 Total = 286.70 Total = 287.41
F= 1.319 F= 1.339 F= 1.342

9.3.4 Stability coefficients by Bishop & Morgenstern (1960)

Bishop & Morgenstern (1960) developed a method using two stability coefficients, m and n,

that satisfy the following equation:

F=m-nr, (9,11)
Whitlow (1990) recalculated the m and n values an extended range and these are given in
over
Table 9.1. The stability number is defined as c'/yH, is the inverse of the definition
which ex-
pressed in Equation 9.4. Using Bishop's simplified method, Chandler & Peiris (1989) pub-
lished results for the coefficients m and n that are in good agreement with the values in Table
9.1. To estimate the factor of safety, the procedure is as follows:
1. Calculatec' / yH from the soil and slope data.
2. For a value of c' / yHjust greater than that found in step 1, use the corresponding sectionof
Table 9.1 and find m and n for na = 1. Use linear interpolation (for dO' values) if necessary,
3. If n is underlined the critical circle is at a greater depth. Use the next higher value of eta to
find a non-underlined n. Use linear interpolation (for O' values) if necessary.
4. Repeat steps 2 and 3 for values ofc'/ )'H just less than that found in step 1.
5. Use Equation 9.11 to obtain two factors of safety for the upper and lower values of c' / )'H.
Calculate the final factor of safety by interpolating between these two values.

Example 9.7

For a slope of 1 vertical to 3 horizontal and of height 17 m, compute the f\177ctor of safety for
\302\260
ru = 0.2. Relevant soil properties are: c' = 12 kPa, \177'= 22 andy= 18 kN/m \177.)))
 Stability of Earth Slopes 377

Table9.I Stability coefficients by Bishop & Morgenstern 1960 recalculated by Whitlow (1990).

c'/yH = 0.000
cot 13 0.5:1 1:t 2:1 3:1 4:1 5:1
lid O\" m 1'\177 m ?l m n m t\177 m II m tl

All 20
\302\260
0.18 0.90 0.36 0.72 0,73 0.90 |.08 1.21 1.45 1.54 1.81 1.88
25 \302\2600.23 1.16 0.47 0.92 0.92 1.16 1.40 1.55 1.86 1.97 2.32 2.41
30
\302\260
0.29 1.43 0.58 1,15 1.15 1.43 1.72 1.91 2.30 2.44 2.88 2.98
35 \302\260
0.35 1.74 0.70 1.39 1.39 1.74 2.10 2.32 2.79 2.97 3,48 3.62
40\302\2600,42 2.09 0.83 1,67 1.67 2.09 2.51 2.79 3,34 3.55 4,18 4.34
c'/ylt
= 0.025

cot 13 0,5:1 1:1 2:1 3:1 4:1 5:1

nd \177b' m n m n m n m n m n m n

1.00 20\302\260 0,52 0.72 0.70 0.76 1.11 1.01 1.53 1.34 1.95 1.69 2,37 2.04
25\302\260 0.59 0,79 0.83 0.96 1.35 1.27 1,87 1.69 2.39 2.13 2.91 2.59
30\302\2600.67 0.88 0.97 1.19 1.60 1,56 2.23 2,07 2+86 2.61 3,41 .3.17
35\302\2600.76 1,00 1.13 1.44 1.87 1.88 2.63 2.50 3.38 3.15 4,14 3.83
40\302\2600.86 1.17 1.30 1.72 2.18 2.24 3.07 2,98 3.95 3.76 4.85 4+56

1.25 20 \302\260
1.00 0.93 1.07 1.02 1.29 1.20 1.60 1.45 1.93 1.76 2.30 2,06

25
\302\260
1.22 1.18 1.31 1.30 1,60 1.53 1,97 1.87 2.42 2+25 2.87 2.65
30 \302\260 1.46 1-47 1.59 1.62 1.95 1.91 2.41 2.33 2.93 2.80 3.49 3,28
35\302\260 1.74 1.76 1.90 1.96 2.32 2.31 2.89 2.83 3.50 3.38 4,17 3.98
40\" 2.04 2,11 2.23 2.35 2.74 2,75 3.43 3,39 4.14 4.04 4.93 4.75
c'/yH
= 0.050

cot 13 0,5:! 1:1 2:1 3:1 4:1 5:1

na CO\" m n \177n n m n m n m n m n

1.00 20 \302\260
0.69 0.78 0.90 0.83 1.37 1.06 1,83 1.38 2.32 /.77 2.77 2.08
25\302\2600.80 0.98 1.05 1.03 1.61 1.33 2.18 1.75 2.77 2,20 3.33 2.64
30\302\2600.91 1.21 1.21 1.24 1.88 1.62 2.56 2.| 5 3.24 2.68 3.91 3.24
35\302\2601.02 1.40 1.37 1.46 2.17 1.95 2.99 2.78 3.58 3,25 4.57 3.96
40\302\2601,14 1.61 1.55 1.71 2.50 2.32 3.44 3.06 4.40 3,91 5.30 4.64
1.25 20
\302\260
1,16 0.98 1.24 1.07 1.50 1.26 1.82 1.48 2.22 1.79 2.63 2.10
25\302\260 1.40 1.23 1.50 1-35 1.81 1.59 2,21 1.89 2.70 2.28 3.19 2.67
30 \302\260 1.65 1.51 1.77 1+66 2.14 1.94 2.63 2.33 3.20 2.81 3.81 3.30
35\302\260 1.93 1.82 2.08 2.00 2.53 2.33 3,10 2.84 3.78 3.39 4.48 4,01
40\302\2602.24 2.16 2.42 2.38 2,94 2.78 3.63 3.38 4.41 4,07 5.22 4.78
1.50 20
\302\260
1.48 1,28 1.55 1.33 1.74 1.49 2.00 1.69 2,33 1.98 2.68 2.27
25\302\260 1.82 1.63 1.90 1.70 2.13 1.89 2.46 2.17 2,85 2.52 3.28 2.88
30\302\2602.18 2.01 2.28 2.09 2.56 2.33 2,95 2.69 3,42 3.10 3,95 3.56
35* 2.57 2.42 2.68 2.52 3,02 2.82 3.50 3.25 4,05 3-75 4.69 4.31
40 \302\260
3.02 2,91 3.16 3.02 3.55 3.37 4.11 3.90 4.77 4.48 5+50 5.12

Solution:

CaIculate the dimensionless parameter c' / yH: 1/ N = 12.0/(18,0 x 17.0)= 0.0392.)))

378 Soil Mechanics.. Basic Concepts and Engineering Applications

c'/\177[H
= 0.075

cot [3 0.5:1 I:I 2:1 3:1 4:1 5:1

na' \177\" m n m n m n m n
\302\260
1 \303\2670020 0.85 0.80 1,09 0.84 1.6! 1.1.__.Q 2.14 1.44 2.66 1.813 3.17 2.13
25\" 0,95 1.01 1.25 1.05 1,86 !.38 2.50 1.813 3.13 2.26 3.74 2.72
30\302\2601.06 1.24 1.42 1.30 2.14 1.69 2.88 .2.20 3.62 2.76 4.36 3.33
35\177 1.19 1.49 1.61 1.56 2.44 2.03 3.31 \177.66 4.18 3,33 5.02 4.00
40'\177 1.33 1.76 1.80 1.82 2.77 2.39 3.78 .3.15 4.79 3.95 5.78 4.7.\177fi6

1.25 20\" 1.34 1.02 1.39 1.09 1.69 1.29 2,07 1,54 2.49 1.82 2.95 2.17
25\302\2601,58 1.28 1.66 1.39 2.00 1.64 2.47 !.96 2,97 2,32 3.52 2.73
30\302\2601.83 t.56 1.94 1.70 2.35 2.01 2,89 2.39 3.50 2.86 4.15 3,36
35\302\2602.11 1.87 2.25 2.03 2.73 2.39 3.36 2.87 4,08 3.46 4.83 4.04
40\302\2602,42 2,21 2.58 2.40 3.15 2,84 3.89 3.43 4.73 4.13 5,60 4.83
1.50 20
\302\260
1,64 1.31 1.71 1,35 I \17792 1.51 2.20 1.73 2.55 1.99 2.93 2.27
25\" 1.98 1.66 2.05 1.71 2.3[ 1.91 2.66 2,20 3.08 2.53 3.55 2.92
30\302\2602.34 2,04 2.43 2.10 2.74 2.36 3.16 2,71 3.66 3.13 4,22 3.59
35\302\2602.74 2.46 2.84 2.54 3.21 2.85 3.71 3.29 4,30 3,79 4.96 4.34
40'\177 3.19 2.93 3.31 3.03 3.74 3.40 4.33 3.93 5.03 4.53 5.79 5,19
c'/ytt
= O. I O0

cot \177 0,5:1 1:1 2:1 3:1 4:1 5:1

m n m n m n m n m n m n

1.00 20\302\2600.98 0.80 1.25 0.86 1.83 1.13 2.41 1.46 2,97 1.83 3.53 2.1\1775
25
\302\260
1.10 1,02 1.41 1.07 2.09 1.4.____\17722,78 1.84 3.36 2.29 4.09 2,72
30 \302\260
1.21 1,25 1.58 1.30 1.72 3.17 .2.25
2,37 3.91 2.80 4.71 3.34
35\302\2601.34 1,50 1.77 1.57 2.08 3.59 2.71
2.68 4.49 3.34 5.39 4.0\1773
40
\302\260
1.48 1,78 1.99 1.87 2.44 4.07
3.01 3.2._\177L 5.10 3.9._.\1777 6,14 4,80

! .25 20 \302\260
1.48 1.03 1.52 1.09 1,86 1,29 2.27 1.55 2.74 1,83 3.23 2.
!.____5__5
25
\177
1.72 1.29 1,79 1,38 2.19 [.63 2.67 1.96 3.21 2.32 3.81 2.7,____\1774
30
\302\260
1 \17799 1.59 2.08 1.73 2.53 2.00 3.09 2.41 3.73 2,84 4\17742 3.35
35 \302\2602.27 1.90 2.40 2,07 2.91 2.41 3.58 2,90 4.30 3.44 5.10 4.04
40\302\2602.58 2.23 2.74 2,44 3.33 2.85 4.09 3.44 4.96 4,11 5.88 4.8.\177__\177

1.50 20 \302\260
1.77 1.30 1.85 1.36 2.07 1.52 2.38 1.73 2,76 2.00 3.14 2.28
25\302\2602.1 ! 1.66 2.20 1.72 2.47 1.93 2.83 2.21 3-28 2.53 3.78 2.9[
30\302\2602.48 2.05 2.58 2.11 2.90 2.38 3-33 2.72 3.86 3.12 4.44 3.59
35\302\2602.88 2,47 2.98 2.54 3.37 2.86 3.88 3.28 4.49 3.78 5.17 4.34
40\302\2603.33 2.94 3.45 3.03 3.90 3-42 4.49 3.92 5.21 4.51 5.99 5.16

For c' I )'H= 0.050,cot[3 = 3.0 and D = t, n is underlined for the range of 0' from 20
\302\260
to 25 \302\260,

thus select D= 1.25 for a deeper critical circle. For 0' =20% m= 1.82, n= 1.48 and for

qb'
= m = 2.21, n = 1.89. Interpolating
25 \302\260, linearly for qb\"
= 22\302\260:

2.21-1.82 \302\260
m = 1.82 + - 20\1770 = 1.976,
\302\260
-
(22.0 \302\260)
(25,0 20,0 \302\260))))
 Stability of Earth Slopes 379

c'/yH
= O, 125

cot [3 0.5:1 I:1 2:[ 3:1 4:1 5:1

1.00 20
\302\260
1.13 0.81 1,43 0,88 2.04 1,1\1775 2.69 1,5\1774 3.26 1.78 3.87 2.12
25\302\260 1.25 1.04 1.60 1.I [ 2.32 1.45 3.06 1.9t 3.74 2.27 4.45 2.72
30\302\260 1.38 1.27 1.77 1.34 2.62 1.7\1778 3.46 2.30 4.25 2.8\177! 5.07 3.37
35 \302\2601.50 1.51 1.96 1.59 2.93 2.12 3.88 2.71 4.82 3.41 5.77 4.05
40\302\2601,61 1.75 2.17 1.89 3.27 2.48 4.36 3._A_8. 5.46 4.0\1776 6.55 \177

1.25 20\302\260 t.64 1,06 1,67 1.t0 2.05 1.32 2.49 1,58 2.98 1.86 3.50 2.17
25\302\260 1.89 t.33 1.94 t.40 2.38 1.67 2.89 1.99 3.48 2.38 4.08 2.75
30\302\2602.t6 1.63 2,23 1.73 2.73 2.04 3.32 2,43 4.01 2.92 4.71 3.41
35\302\2602.45 1.95 2.56 2.09 3.11 2.45 3.80 2.93 4.59 ,3.50 5.41 \177
40\302\2602,77 2.30 2,92 2.49 3.54 2.91 4.33 3.49 5,24 4.16 6.21 \177
1.50 20
\302\260
1.92 t.32 2.02 t.39 2.23 1.55 2.57 1.75 2.96 2.00 3.40 2.29
25\302\2602.26 1.68 2,37 1.75 2.64 1.97 3.03 2,23 3.50 2.55 4.02 2.91
30\302\2602.63 2.07 2.75 2.15 3.07 2.43 3.53 2.75 4.08 3.15 4.69 3.60
35\302\2603.04 2.50 3.t6 2.58 3.55 2.92 4.08 3.32 4.73 3.81 5.44 4.36
40\302\2603.50 2.98 3.63 3.07 4.09 3.49 4,7 t 3.98 5.46 4.57 6.28 5.23
=
c'/'\302\242H O. 150

cot [3 0.5:1 1:1 2:1 3:1 4:t 5:1

nd \177\" m n m n m n rn n m n rn n

1.00 20 \302\260
t .25 0.8 t 1.58 0.89 2.25 I.
1\1776 2.89 1.4\1774 3.57 1.8\1770 4.21 2.1_...\1775
25
=
1.37 t.02 1.75 1.12 2.53 t.80 4.01
t,45 3.24 2.27 4.78 2.77
30 \302\260t.50 t.25 t.93 1.36 2.83 t.78 3.64 2.24 4.54 2.79 5.41 3.39
35 \302\260t.65 t.53 2.12 1.6t 3.t4 2.14 4.09 2.7t 5.10 3.38 6.09 4.09
40\302\260t.80 t.82 2.33 1.89 3.49 2.53 4.57 3.24 5.74 4.05 6.86 4\17785

1.25 20 \302\260
1.79 1.07 t.80 1.I0 2.22 t.32 2.69 1.59 3.22 t.86 3.77 2.17
25\302\2602\17703 t,33 2\17707 1.40 2,55 t.68 3.09 2.0t 3.71 2.37 4.33 2.76
30\302\2602.30 1.63 2.37 1.74 2.90 2.06 3.51 2.44 4.22 2.92 4.96 3.38
35\302\2602.60 1.96 2.69 2.08 3.28 2.47 4.00 2.94 4.81 3.50 5.66 4.10
40 \302\2602\17792 2,33 3\17705 2.44 3.72 2.92 4.53 3,48 5.46 4,17 6.44 &92

t,50 20 \302\2602.05 1,33 2,15 1.39 2,38 t.54 2.74 t.75 3.t5 2.0t 3.63 2,30
25\302\2602.39 1.68 2.51 1.76 2.77 t.97 3.19 2.23 3.67 2.55 4.23 2.90
30\302\2602.76 2.07 2.89 2.16 3.22 2.43 3,70 2.75 4.26 3.14 4.90 3.57
35\302\2603.16 2.50 3.30 2.59 3.69 2.92 4.24 3.3 t 4.90 3.79 5.64 4.33
40\302\2603.62 2.98 3.76 3.07 4.23 3.48 4.87 3,95 5.63 4.54 6.47 5.t9

1.89-1.48
n = 1.48 + (22.0 \302\260- = 1.644,
20.0 \302\260) F\177
= 1.976 -1.644 0.2 = 1.647.
\303\227
\302\260
(25.0 - 20.0 \302\260)
For c' / yH = 0.250, cot[\177 = 3.0 and D = 1, for q\177'= 20 \302\260, m = 1.53, n = 1.34 and for 0' = 25\302\260,

m = 1.87, n = 1.69. Interpolating linearly for 0' = 22\302\260:

1.87 -1.53
m = 1.53+ \302\260
(22.0\302\260
- 20.0 =
\302\260)1.666,
- 20.0
(25.0 \302\260))))
380 Soil Mechanics: Basic Concepts and Engineering Applications

1.69- 1.34
n = 1.34 + (22\"00
-
20\"00)= 1.480,F2 = 1.666- 1.480x0.2 = 1.370.
(25.0- \302\260
20.0 \302\260)

1.647- 1.370
F = 1.370-+ -
0.025) = 1.53.
(0.050-
Find Fby interpolation: (0.0392
0.025)

9.3.5 Deterraination of pore pressure ratio ru

One of the major factors causing instability and subsequent failure of slopes is the increase of
the pore pressure due to natural causes or construction. An increase in pore pressure reduces
the effective normal stress on the potential failure plane, which in turn reduces the shear
strength on this surface. Values of ru used in design must represent thosestates that are created
during and at the end of construction, and later in the lifetime of the slope. In the construction
process of an embankment, pore pressures are increased above the original level as a result of
loading. Depending on the type of the material, an undrained condition may prevail until the
excess pore pressures dissipate over time. An estimate of ru can be made by utilising Equation
4.26. Consider ui to be the initial pore pressure at the base of slice. As a result of the loading
equivalent to the weight of the slice, the pore pressure increases to ui + yhB, where B is the
pore pressure coefficient obtained in the laboratory triaxial tests that simulate field conditions.
The design value of r, may therefore be taken as:

ru = (u i +Ue)/Yh
= (u
i +\302\245hB)/yh
= u
i/yh+ B (9.12)
In earth dams subjected to an increase in the height of the external water in the upstream side,
the pore pressure will increase and may make the downstream side unstable. To estimate the ru

vales, it is necessary to construct the corresponding flow net representing the seepage through
the earth dam from the upstream side to the downstream side. Any rapid drawdown in the up-

stream side will eliminate the contribution of the external water pressure and may cause the
failure of that side. At the same time, the effective stresseswill increase due to a decrease in
pore pressure which is equal to the change in the total vertical stress ywhw times the pore pres-
sure coefficient B. Therefore, the new r, value is:
r, = (u i -Ue )/Yh
=
(Ywhw
-
YwhwB)l \177h
=
(\177hwlyh )(l- B), thus:

r,, = (u i / 7h)(l
- B) (9.13)
Trial circles

WT _/.\1777\177 Equipotentia I

Figure 9.6. Determination of pore pressure ratio at steady flow.)))

 Stability of Earth Slopes 381

In general the change in the level of the external

with water, the flow net within the embank-
ment will
change and it is necessary to search for a critical flow net to compute the corre-
sponding pore pressures. An example of flow net is shown in Figure 9.6 for the case where the
water moves from the upstream side to the downstream side. In the case of a cut slope, pore
pressures will decrease but, depending on the drainage conditions, will recover over time and a
new equi/ibrium will be achieved. The evaluation of slope stability in undrained conditions
and the application of the effective stress analysis with a critical ru value may closely bracket
the above different cases.
The determination of an average value of r, for a specified trial circle can be justified by a
mathematical model. Bromhead (1992) suggestssystematic calculations to define an average
value for ru. According to this method, the soil mass between the slope boundaries and the
hard stratum underneath is divided into vertical strips of equal width. Each strip is then di-
vided vertically into three equal portions and ru is determined at the centroid of each port/on.
The averageof the ru values of the middle portions (weighted by the area of each portion)
represents, the average ru for a general failure. For shallow failures, the average of the upper
portions and for deep failures the average of the lower portions can be used.

9.4 STABILITY OF INFINITELY LONG EARTH SLOPES

9.4.1 Plantar failure mechanism

A two dimensional infinitely long earth slope is shown in Figure 9.7. For the purpose of analy-
sis it is assumed that the material is homogeneous. The failure mechanism to be a
is assumed
plane parallel to the ground surface on which the \177ormal and shear stresses are yHcos2[3
and

yHsin\1773cos[3 respectively as was shown in Section 8.3.3 for a rhombic element. In practice, this
mechanism is applied to the case when a soft material (e.g. clay) of very long length with con-
stant slope may slide on a hard material (e.g. rock) having the same slope. The objective of the
analysis is to determine either the critical slope or the critical height or, alternatively, the factor
of safety based on a concept similar to that of the circular failure surface described previously.
Steady seepage conditions in the slope will reduce the factor of safety.

Figure 9.7. Planar failure mechanism for an infinite slope.)))

382 Soil Mechanics: Basic Concepts and Engineering Applications

(a)

Figure 9.8. State of stress in an infinitely long slope:(a)ca,Cu = 0 soil, (b) c', \302\242\"
soil.

9.4.2 earth = 0 soil

Stability of infinitely long slopes in cu, d\177u

On any plane parallel to the ground surface the ratio of the shear stress to the normal stress is
constant and equal to tanl3. Thus, the states of stress on these pianos are located on the line OF
in the o, coordinate
\342\200\242 system, as shown in Figure 9.8(a). If the normat stress is less than OC,

(point T on line OF) the shear stress wilt be Iess than cu and the factor of safety is:

F = SF'/ST, F = cu/(yH sinl3cosI3) (9.14)

For normal stresses greater than OC the material will (theoretically) fail. The corresponding
critical height Hc for a specified slope angle 13 is found by setting F to I in Equation 9.14:

Hc = c, /ysin 13cos\177
= 2c /ysin(213)
u (9.15)

Similarly, a criticat slope angle 13c may be defined for a specified value of H as follows:
=
\177c 0\"5sin-l(2cu/yH) (9.16)

9.4.3 - no
Stability of infinitely tong earth slopes in c', d\177'soil seepage
The states of stress on planes parallel to the ground surface and with 13 > \177b\"
is represented by
line OF in Figure 9.8(b). Defining the factor of safety as the ratio of SF' (shearstrength) to ST

(shear stress), then:

SF' c'+OStan\177b\" c' tan
F '
OS tan[3 2
ST yH cos [3tan[3 tan[3
t
C tan
F= + (9.17))))
sin
\"\177'H 13cos13 tanl3
 Stability of Earth Slopes 383

As the depth of the sliding plane increases, the normal stress on the plane approaches its limit-

ing value OC. The critical height Hc is defined by setting Equation 9.17 to unity:
\342\200\242 2
_ c sec 13
H e ( (9.18)
--\177 tan13_tan\177,)

For the case where 13 < q\177',the factor of safety is always greater than 1 and is computed from
Equation 9.17. This means that there is no limiting value for H, and at an infinite depth the
factor of safety approaches:
F = tan 0\"/tan13 (9.19)

For a granular material with c' = 0 and 13 < \302\242',

the factor of safety is computed from Equation
9.17 (or 9.19), The case where 13 > q\177'and c' = 0 is always unstable aad cannot be applied to
practical situations. This means that the critical value of the slope angle is:

13c
= 0' (9.20)
Example 9.8

An infinitely long slope is resting on a shale formation with the same inclination. The height
of the slope is 3.2 m. Determinethe factor of safety, the shear stress developed on the sliding
surface and the crtucal height. [3 = 25 , g = 17.5kNlm , c = 12 kPa, and \177
= 20 .

Solution:
\302\260
Using Equation 9.17: F = 12.0/(17.5\303\227 sin
3.2\303\227 25.0\302\260cos25.0\302\260)
+ tan 20.0\302\260/tan 25.0 = 1.34.

z = yH sin 13cos 13
= 17.5 \303\227 sin 25
3.2 \303\227
\302\260
cos
\303\227 25,0
\302\260
= 21.4 kPa.
The critical height is calculated from Equation 9.18:
H e = (I2.0\303\227
sec
2
25,0 \302\260
/ 17.5)/(tan 25.0
\302\260
- tan 20.0 \302\260)
= 8.16 m.

' - steady state

9.4.4 Stability of infinitely long earth slopes in c', \177p soil seepage
The of steady state flow
case parallel to the slope angle 13 and with the water table at the

ground surface is shown in Figure 9.9(a). From the geometry of the flow net, the pore pressure
at point M of the base of an arbitrary vertical element is given by:

u = MN \303\227Yw= = 2
cos [3yw
MV \303\227cos13\303\227y\177,H

'
The factor of safety is: F -- -
_- \"rf c\" + cy'tan0' - c'+(yHcos213-YwHCOS2\177)tan(\177
\342\200\242
yH sin 13cosl3 yH sin 13cos13
- c\"
+ tanq\177' Yw tan0\"
F (9.21)
yH sin13cos[3 tan13 ytan 13

The first two terms of Equation 9,21 are identical to Equation 9.17; the third term indicates the
reduction in the factor of safety due to steady state flow. If the ground water level is at some

depth but parallel to the ground surface, Equation 9.21 is modified as follows:)))
384 Soil Mechanics: Basic Concepts and Engineering Applications

\177
Equipotential. N\177 Equipotential v N
\177.\177-\"\177\"

ines

(a)

Figure 9.9. Seepage in an infinitely long slope with water at the ground surface\342\200\242

c' tanO' y\177vhw tan0'

F = + (9.22)
\177'H sin 13cosl3 tanl3 \177'H tan 13

where hw is the height of the water above the base of the slice. Substituting ru for \"fwh,,. / yH

(Equation 9.8) into Equation 9.22 we obtain:

c\" tan0\" tanO'

F = \177
ru =m-nru (9.23)
\177,Hsin[3cos[3 tant3 tan[3
which is the fundamental relationship of the Bishop and Morgenstem approach to the stabiIity
of shallow slopes. If the unit weights of the saturated zone and the zone above the water table
are not the same, the term \177'H in Equations 9.22 and 9.23 must be replaced by Z'\177H. Equation
9.21 may be conveniently presented in the following form:

F= \177' tan\177'
\177- (9.24)
\"gHsint3cos[3 5' tanl3

By setting Equation 9.24 to unity the critical height is calculated as:

2
- c\177sec \177
H e (9.25)
y tan\177- 5\"tang 0\"

For the values of tanl3 < (g' / \177')tan\177', the factor of safety expressed by Equation 9.24 is always
greater than 1.0. At infinite depth the factor of safety is given by:

F = \177/tan\177'/\177'tan[\177 (9.26)

From Equation 9.26 we can also calculatethe factor of safety for a granular material with c\" =

0. In this case tan[\177 must be less than (y' I \"\177)tan\177', otherwise the slope will not be stable. By
setting Equation 9.26 to unity a critical slope angle is defined for granular materials:

= tan-1 tan
13 c (\302\245, O'! \177) (9.27)

The case where the steady flow is not parallel to the ground surface is shown in Figure 9.9(b).
It can be shown that the pore pressure at point M is defined by:)))
 Stability of Earth Slopes 385
YwH
u = YwMN = (9.28)
1+ tan o. tan 13

where is
o\177 the angle of the flow lines to the horizontal. The corresponding factor of safety is

derived in a similar way to that undertaken for Equation 9.21:

c\" tan\177\" yw tan\177'(l + tan21\177)

F = + (9.29)
7Hsin[\177cos[3 tanl8 ytanl\177(l+tano\177tanl8)

The critical height for a c', \177b\"soil and critical slope angle for c*= 0, \302\242/soil are as follows:
2
- C'See \177
H
e (9,30)
+ tan213
y[tan[3-tan qS'(1
1 +1
\302\245w\302\245 )]tanot tan 13

Yu' l+tan2[\177c
13c=tan-l[(1 )tan\177b'l (9.31)
1 +
\302\245 tan o\177tanl\177c

It is seen that Equation 9.31 must be solved using iteration.

Example 9.9

A long slope of H = 4.5 m and 13

= 22 \302\260
be constructed of material
is to having the following
properties: =
Ysa\177 20 kN/m 3, yd .ry = 17.5 kN/m 3, c' = 10kPa, and \302\242\177\"
= 32 \302\260.
Determine the factor
of safety when: (a) the slope is dry, (b) there is steady state seepage parallel to the surface with
the water level 2 m above the base, and (c) the water level is at the ground surface.

Solution:

\302\260
10.0 tan 32.0 -
(a) From Equation 9,17: F = \177-- 1.91.
17.5x4.5 xsin cos22.0
\303\227
22.0\302\260
\302\260
tan
\302\260
22,0

(b) Use Equation 9.22 and replacing yH with ZyH:

10.0 tan 32.0 \302\260 9.81 x 2.0x tan 32.0 \302\260

F= -I
\302\260 \302\260 \302\260\"
(17.5 x2.5 + 20.0x 2.0)sin 22.0\302\260cos22.0 tan 22.0 (17.5\303\2272.5 + 20.0x2.0) tan22.0

F = 1.89 - 0.36= 1.53.(c) Use Equation 9.22:

\302\260 \302\260
10.0 tan 32.0 9.8 lx4.5x tan32.0
F=
\302\260
+ =1.11.
\302\260
\302\260 \302\260
(20.0x 4.5) sin 22,0 cos 22.0 tan 22.0 (20.0x4.5) tan 22.0

9.5 STABILITY OF REINFORCED AND NAILED EARTH SLOPES

9.5.1 Circular failure surfaces for slopes reinforced or nailed by geosynthetics

Earth slopes can be stabilized by geosynthetic reinforcement with various deployment

schemes and spacing (Schaefer et al., 1997). Examples of applications of geosynthetics in geo-
technical engineering may be found in Koerner (1991) and Ingold (1994). Normally the rein-)))
386 Soil Mechanics: Basic Concepts and Engineering Applications

forcement is placed horizontally with uniform vertical spacing. However, reduced vertical
spacing may be used in the deeper regions where there are higher lateral stresses due to grav-
ity. If the foundation underneath is a weak saturated soil, the reinforcement may be applied
only at the base of the slope. The length of the reinforcement must extend by the length tr be-

yond the failure surface as described in Section 8.8.2. Its vertical spacing is designedby taking
into account the thickness of the compacted layers and the weight of the soil above the rein-

forcement that provides frictional resistance for the embedded length lr. For a reinforced slope,
the concept of a circular sliding surface may be applied by modifying Fellenius' or Bishop's
simplified methods to account for the moment contributions of the tensile forcesdeveloped in
the layers of reinforcement. The problem becomes statically determinate by making appropri-
ate assumptions, which include a linear distribution of the lateral stresses resisted by rein-
forcement. In the case of parallel reinforcement in a cu, 0u = 0 soil, adding the moments of the
geosynthetic tensile forces to the resisting moments results in the following equation:

-
CuLaR + \177[T(y C y)]j
F - j=l (9.32)
Wd

where Tj is the geosynthetic yj is the vertical distance of the reinforcement

tensile force, from
the x-axis (Figure 9.10(a)),and number of the geosynthetic
m is the layers. Note that by in-
cluding the additional moment in the denominator of Equation 9.32, the tensile force required
for equilibrium is Tj/F. In the presence of a surcharge load on the upper ground surface and
external water pressure at the front of the slope, the corresponding additional moments must
be taken into account. For the slope in the c', 4' soil shown in Figure 9.10(a), Bishop's simpli-
fied method is modified as follows:

-
\177 {\177cb+w(1-r\177)tano'l/ma}i Y)]j
i=l +\177 =
j\177lfT(Yc
F= (9.33)
i=n

\177(wsina)i

Area = 1 12 yH2k

(a)

Figure 9.10. Parallel geosynthetic reinforcement.)))

 Stability of Earth Slopes 387

The total reinforcement force T\177o,al is the sum of the tensile forces in the reinforcement; it is
equivalent to the integral of the lateral soil pressure (rt\177with the pressure coefficient k less than
ka (for F = 1) due to the slope angle. It is convenient to assume a linear distribution for lateral
stress with the depth as suggestedby 9.33 can be solvedfor a speci-
Jewell (1991). Equation
fied F to yield Ttotal, The procedure value of Tto\177al, corresponding
is repeated until a maximum
to the critical circle, is obtained.This type of analysis may overestimate the reinforcement
forces and the current trend is to assume that the fult strength of the soil on the critical circle is
mobilized (F = 1).A final check on the overall stability must be carried out using trial circles
that are deep and well beyond the reinforcement. This will eliminate the possibility of stiding
where the foundation soil has poor strength properties. In order to evaluate the tensile forces in
the reinforcement, the following procedure should be followed:
1. Specify a factor of safety F not less than 1.4.
2. Substitute F into Equation 9.33 and compute the sum of the moments contributed by the re-
inforcement: AM =Z[T(yc-y)]j for the corresponding trial circle.
a linear distribution -
3. Assume for Crh in the form of: \177 =
\302\245(H y)k, thus Trota l =yH2k/2.

Equate the moment of the resultant of the linear distribution to the sum of the tnoments con-
tributed by the reinforcement (step 2) and compute Ttotal and k. Repeat the procedure for a
number of trial circles until a maxi\177num value for Ttotal (or k) iS obtained. Ttotal is located H / 3
above the base of the slope. Note that for F > 1, the magnitude of k may be greater than ka.

- =
AM = \177_\177[T(Yc Y)]j T\177orat(Yc
-H /3)=(yH2k /2)(Yc -H /3) (9.34)
j=i
4. Divide the lateral stress distribution into horizontal strips, The area of each strip is equal to
the tensile force of the reinforcement (per metre run of the slope) as shown in Figure 9.10(b),
5. If a surfaceIoading q exists:
=
[y(H
- y) + q]k, =
yH 2k / 2 + qHk and:
ot\177 t
T\177o\177a

i=tt\177

AM =
[T(y c
- : (\177'H *-k / 2)(y C
- H / 3) + qHk(y C H / 2) (9.35)
\177 y)\177
i=l

6. Calculate the required reinforcement length according to Section 8.8.2.

In the construction of Equations 9.32 and 9.33, the reinforcement contribution in penetrat-
ing the faiIure surface has been ignored. the shear
strength parameters and the factor of
Thus,

safety may be multiplied by factors greater than 1 (Koenaer & Robins, t986).
Soil nailing is applied in cut slopes where rods are inserted into the slope as excavation pro-
ceeds.Using the method suggested by Koerner (1984), it is possible to construct fill sIopes
with geosynthetic surface-deployed reinforcement (Figure
nailing. This is a 9.11(a)) that is
nailed into the
slope as the
proceeds, construction
Assuming that each slice contains a rein-
forcement element, Bishop's simplified method can be modified to take into account the mo-
ment and vertical force equilibrium contributions of the reinforcement element as follows:

trta
F = i=1 I. \177i
(9.36)

,\177_,(wsino:- Td / R) i
r=l)))
388 Soil Mechanics: BasicConcepts and Engineering Applications

Trial circle

(a) (b)
Figure 9.1 I. Application of the slice method to a slope nailed by geosynthetics.

where c0i is the angle of the reinforcement from the vertical and di is the distance of the rein-
forcement from the centre equation of the trial circle. This may be modified to include the con-
tributions of nails surface. One type of modification
penetrating the failure
suggested by Ko-
erner & Robins (1986) is to replace F with F/ (1+39where f> 0, and use improved shear
strength parameters c,n >_ c' and 0m >- q'- The tensileforce at each layer of reinforcement is
expressedin terms of its horizontal projection Tit,. Each horizontal projection is then expressed
in terms of the (unknown) lateral soil pressure coefficient k, obtained by estimating a reason-
able width for the stress strip shown in Figure 9.10(b). The maximum value of k is determined
by searching for a critical sliding circle. The horizontal component of each reinforcement
force can then be found which in turn is used to compute Ti.
Example 9.10
A 7.2 m high slope, of 1.0 horizontal
which has to 1.8 vertical, is to be reinforced
a batter with
3
horizontal geosynthetic of the soil are: c' = 0, 0' = 35\302\260,
elements. Properties \177/= 19 kNhn and
= 0.4. Two trial toe circles are to be considered, with xc, Yc coordinates
ru (0, 13.60 m) and (0,
8.04 m). For both circles compute the totaI tensile force in the reinforcement assuming a factor
of safety of 1.4 and using Bishop's simplified method. Relevant details of the geometry are
given in the first three columns of the following two tables.

Solution:
For the trial circle with R = 13.6 m using Equation 9.33:
F = 1.4 = (340.2+AM/13.6)/371.0 -\177 AM = 2437.12 k_N.m.
From Equation 9.34:

AM = 2437.12 = T\177o,at (13.6- 7.2 / 3) --\177

Ttora I
= 2437.12111.2 = 217.6 kN.
For R = 8.04 m, F = 1.4= (209.9
+ AM / 8.04)/275.8 --\177 AM = t416.81 kN.m, and

AM = 1416.81 = (8.04
- 7.2/3) -\177 = 1416.81/5.64 = 251.2 kN > 217.6 kN.)))
Ttora I Ttora I
 Stability of Earth Slopes 389

R = 13.6 m, n = 6, b = 2.0 m for all slices, F=1.4, r,= 0.4\177

h o\302\242 w \177in\177 w(l

- / ma
ru)tan\177'
Slice
(m) (deg.) (kN) (kN) (kN)

I 1.7 4.2 64.6 4,7 26.2

2 5.0 IZ7 190.0 41.8 73.5

3 6.2 21.6 235.6 86.7 88.9

4 5.2 31.0 197.6 101.8 74.5
5 3.7 41.4 140.6 93.0 54.6
6 1.4 54.0 53.2 43.0 22.5
Total: 371.0 340.2

R= 8.04m, n = 4, b = 2.0 m for all slices, F= 1.4,r\177,= 0.4.

h w w( I
-
c\177 wsinc\177 r\177)tan\177' / m,\177
Slice
(m) (deg.) (kN) (kN) (kN)

l 1.7 7.1 64.6 8.0 25.7

2 4.7 21.9 178.6 66.6 67.3
3 5.3 38.4 201.4 125.1 77.3
4 2.3 60.5 87.4 76.I 39.6
Total: 275.8 209.9

9.5,2 Application of a two-part wedge mechanism to slopesreinforced by geosynthetics

The stabiIity of a reinforced slope may be assessed by the two-part wedge mechanism shown
in Figure 9.12(a). With a verticalinter-wedge boundary, the mechanism is defined by three
dependent variables h, 0t and 02. If the inter-wedge boundary is not vertical then an additional
variable is needed to specify the boundary. In the evaluation of stability, only force equilib-
rium is used. The faiIure criterion is assumed to apply on the three sliding surfaces, which in
turn implies that the shear strength on these surfaces is fully mobilized. In the two-part wedge
method adopted by the Department of Transport, UK (1994),it is assumed that the friction an-

gle on the inter-wedge sliding surface is zero; which results in reduced computational effort. A

base-sliding factor to modify the of

\177b' the lower wedge may also be applied. Free body dia-

grams of the wedges are shown in Figure 9.12(b) where the total reinforcement force TtotaI is

Figure 9.12. Application of a two-part wedge analysis for slopes reinforced by geosynthetics.)))
390 Soil Mecl\177anics: Basic Concepts and Engineering Applications

the sum of Tland T2 corresponding to wedges 1 and 2 respectively. The force system is stati-

cally determinate and can

T\177o\177at be formulated by considering horizontal and vertical equilib-
rium of both wedges:

W 1 (tan - tan q\177')+ (U tan 9'- CI ) sec01

01 1
T\177otal
= +
1+ tan 0 t tan 0'

- tan - C
W2 (tan 02 + (U
\302\242') 2 tan \302\242' 2) sec 02
(9.37)
1+ tan02
tan 9'

where WI, weights of the wedges

W2 are 1 and 2, Ct, C2 are the forces due to cohesion acting
on thesliding bases, and UI, U2 are the forces due to water pressure acting on the sliding
bases. The resultant of any surcharge loading acting on the slope surface or upper ground sur-
face must be added to W1 or W2 where appropriate. By optimising (usually by iteration) the
three variables of the mechanism, the critical two-part wedge corresponding to maximum
Ttotal can be obtained. Similar to the circular method, the computed value of T\177otat is assumed
to be [inearly distributed along the slope height. Jewellet al. (1985) and Jewell (1991)used a
two-part wedge mechanism (as well as a log-spiral failure surface) to estimate the soil pressure
coefficient soil with different values ofr,. Dependingon the height
k for a c\" = O, \302\242\" of the wa-
ter in the slope, an expression for an average value ofru was defined. The k values so obtained
were nearly identical to those obtained from circular failure surfaces. A typical chart for ru = 0
is shown in Figure 9.13.

Example 9.11
For the two-part wedge mechanism shown in Figure 9.14, calculate the total force in the hori-
zontal reinforcement and equivalent soil pressure coefficientk. H = 10m, 13
= 60 \302\260,=
y 19
kN/m 3, c' = 0 kPa, and = 35 \302\260.
\302\242\"

\177 0.6
15\"
\302\242'=
\342\200\242
0.5
e' = 20\"
\"\177 0.4 o
,0 \302\242'=25
,C,
\177 0.3 \177'
= 30 \302\260
\302\260
\177
35
\302\242'=
0.2
= 40 \302\260
\302\242'

\177- 021

m 0
30 40 50 60 7O 80
I\177(degrees)

Figure 9.13. Variation of k wi\177h slope angle for

I\177 c\" = 0, \302\242'
soil and r, = 0
(Exxon chemicals, t 992).)))
 Stability of Earth Slopes 391
D C
\"\177-
3.3m
i
/\177 I\177E=WT
H = 10 60\302\260

7\17701..\177

Figure 9.14, Example 9.1I.

Solution:

From the geometry of the slope in Figure 9.14: AD = 11.55m, AB = 6.67 m, BD = 6.67m, DC
= 3.85 m, BF (normal to AD) = 3.33 m, BE =3.89m, he = depth of water at point B = 3.37 m,
and hA =
depth of water at point A
= 6.7 m. Note that BC is parallel to AD.
W1 -\177
(BDxDCxl.O)y/2
= 6.67 x 3.85x 19.0/2 = 243.9kN.
W
2
= (ADxBFxI.O)y/2 = 11.55x3.33x19.0/2 = 365.4 kN.

u B =\177whB -9.81x3.37 =33.0 kPa, u A =\177(wha =9.81x6.7

= 65.7 kPa.

U\177
= (BExu
B xl.0)/2
= 3.89x33.0/2 = 64.2kN.
U
2
=
AB(uB + u A )1.0/2
- 6.67(33.0+ 65.7)/2 = 329.2kN. Usi\177lg Equation 9.37:
\302\260 \302\260
Tto\177at
= [243.9(tan 60.0 tan 35.0 \302\260)
+ (64.2 tan 35.0 \302\260)
sec 60.0\302\2601/(1 + tan 60.0 tan 35.0 \302\260)
+

[365.4(tan 30.0 \302\260

- tan 35.0 +
(329.2 tan 35,0 \302\260)
\302\260) + tan 30.0
sec 30.0 \302\260]/(1
\302\260 =
tan 35.0 \302\260)

154.4+ 157.6 = 312.0kN. Ttota ! = 312.0 = \"\177H2k/2 = 19.0x 100\303\227k/2 \177, k = 0.33.

9.5.3 Slopes coasrructedon soft clay

Soft clay is a soil with a significant volume change beyond the preconsolidation pressure
(Ladd, 1991), Normally and lightly consolidated clays with c,, < 30 kPa are consideredto be
soft (Jewell, 1996). If the foundation soil underneath a slope with no internal reinforcement is
soft clay, geosynthetic reinforcement is placed on the interface of the two soils to increase sta-
bility against possible horizontal sliding, In a c' = 0, \302\242Y
fill, the failure mechanism shown in

Figure 9.15(a) may be used to against sliding. The lateral

assess the stability active thrust Pa
on the vertical inter-wedge boundary BE is resisted by the shear stresses developed on the re-

inforcement along AB. Shear stressesare mobilized by the weight of the triangular block ABE
and the maximum value of their resultant is: Wf\177tandp'. where W is the weight of the block

ABE, andf\177 is defined by Equation 8.76. The factor of safety against sliding is Wf\177tanO'/Pa. To

calculate Pa it is convenient to assume the full active state on the vertical boundary BE. This
model implies that no shear stress is transferred to the surface of the soft clay. However, de-
pending on the extensibility of the reinforcement, a part of the shear force may transfer to the
surface of the soft clay. This will reduce the bearing capacity of the soft clay (Chapter 10) and
may cause horizontal sliding of the soft clay, particularly if it has a finite thickness.)))
392 SoilMechanics:BasicConcepts and Engineering Applications

W E C D wrlE

Pa
', f\177cCuB'\177- ',
Reinforcement . t Active
Soft clay Passive \177 Soft clay I

-\17711\177.--\177 , Cu B

(b)
Figure Outward
9.15\303\267 sliding of a slope on soft soil.

The proportion of the shear force so transferred is estimated from the properties of the rein-
forcement and empirical equations suggested by the relevant codes. The length of the embed-
ment of the reinforcement beyond point B is calculated in usual way.
\302\260 -
For cases where 13 < 30 the use of a two-part wedge analysis is not recommended and the
force in the reinforcement the equivalent k may be calculated using
and the circular failure sur-
face method. The can be extended to the case where the soft clay has a finite
solution thick-
ness D and is resting on a hard horizontal stratum (Figure 9.15(b)). Active and passive states
are created over the boundaries BB' and AA' respectively. On the boundary AB, which is the
interface of the soft soil and the underside of the reinforcement, the resisting shear stress isfbc
cu, where fbc must be determined by a modified
\303\227 direct shear test. On the lower boundary
A'B' full mobilization of undrained cohesion may be assumed. Knowing all forces applied to
the sliding mechanism, an estimate of the factor of safety against sliding can be made. The
force in the reinforcement will be the sum of shear forces acting on both surfaces of rein-
forcement. Note that the calculation is carried out for a unit thickness of the slope and this
must be taken into account if strip-type reinforcement is used. Partial factors of safety may be
applied to the strength properties of slope fill and soft clay. These are of the order of 1.2 for
drained conditions and 1.5 for undrained conditions (Jewell, 1996). The soft clay may be in
undrained, partially drained or drained co\177lditions, the latter occurring after a long time under
load. The undrained foundation the most critical condition.To estimate
is probably the elastic
and consolidation settlement of the soft clay,
an estimate of the contact pressure on the surface
of the soft clay must first be made. One (versatile) approachis to use the linear contact pres-
sure conceptunder the length AB caused by the weight of the triangular block of ABE and the
horizontal force Pa- Making use of Equation 5.98:

= \",/'H 2
qL = \177lHka
= - k tan = tan > 0
qmax qe (1 a 213 ) , qmin 13 (9.38)

Example 9.12
A fill slope of H = 6 m, [3
= 25 \302\260,
c\" = 0, \302\242\177'=35
and 7
\302\260,
= 19 constructed on a soft clay
kN/m 3, is
soil with geotextile reinforcement on the horizontal interface soils. From laboratory
of two test

results, fb
= 0.7 and fbc = 0.6. The thickness of the soft clay is 3 m, c, = 15 kPa and y = 17)))
 Stability of Earth Slopes 393

kN/m 3. Determine: (a) the factor of safety against outward sliding of the side of the slope, (b)
the factor of safety against outward sliding of the soft soil, (c) the tensite force in the rein-
forcement, and (d) the contact pressure at points A and B of Figure 9.15(b).

Solution:

(a) WABE
= H \303\227
Hcot =
13\303\2271.0\303\227\177/126.0\303\2276.0cot25.0\302\260\303\2271.0\303\22719,012
= 733.4 kN.

Using Equation 8.6: ka = (1 - sin 35.0 + sin

\302\260)/(1
= 0.271,
35.0 \302\260) and from Equation 8.10:

Pa(fitl)
= ?H 2k
a/2 = 19,0\303\2276.02 \303\2270.271/2 = 92.7 kN.

Pzl
=
Wfl\177 tan\177\"
= 733.4\303\2270.7\303\227
tan35
\302\260
= 359.5, Fs = 359,5/92.7= 3.9,
(b) The active force acting along BB\" is in accordance with F\177uation 8.10 with ka = 1:
Pa =0'5\177 'H2 +qH
= 0.5\303\22717.0\303\2273.02 +(19.0\303\2276.0)\303\2273.0-2\303\22715.0\303\2273.0=328.5kPa.
-2cull
The passive force on the boundary AA\" is calculated from Equation 8.11:
= 0.5 \303\22717.0\303\2273.02+ 2\303\22715.0\303\2273.0= 166.5 k.Pa.
-Pp
The shear force developedbeneath the reinforcement is:
= \302\260
B\303\227t.0 = 0.6\303\227
15.0x6.0cot 25.0 = 115.8 kN.
P\1772 f\177cc\177

On the sliding surface between the soft clay and the hard stratum, the mobilized shearforce is:
\302\260
= = 15.0
cu B \303\227 of the block:
1.0 6.0
\303\227 cot 25 = 193.0 k_N. The factor of safety against
Pt3 sliding

FS = (166.5
+ 115.8 + 193.0) / 328.5 = 1.45.
(c)The total tensile force in the reinforcement is: Pa (flit) + Pt:\177 = 92.7 + 115.8 = 208.5kN.
(d) From Equation 9.38:
- 0.271 :z
qrnax
= 19.0 \303\227
6.0(1 tan
\303\227 25.0 o) = 107.3 kPa, (right edge).
= tan-
19.0\303\2276.0\303\2270.271\303\227 25.0 = 6.7 kPa, (left edge).
qmin

9.6 GENERAL SLOPE STABILITY ANALYSIS

9.6.1 Principles of a general slopestabitfty analysis

The conventional limit methods described above do not satisfy

equilibrium all the conditions
of equilibrium. Fellenius' assumes the tangential
method and normal forces on the right and
the left of a slice are equal and have identical points of application, Bishop's simplified method
assumes the tangential forces on both sides of a slice are equal. A general stability analysis
must satisfy all conditions of equilibrium and brief descriptionsof several general methods
follow,

9.6.2 Bishop'smethod
Referring to Figure 9.5(b), vertical and horizontal equilibrium of each slice gives:
AX = X
l
- X
2
= N'(cos + si\177\177tan
c\177 \177b'/F) + c'b tan ct/F - w(l -r u )

/kE = - E = N'(cosc\177tan 0' / F

- si n o\177)+ c'b / F -
E\177 2 wr
u tan c\177)))
394 Soil Mechanics: Basic Concepts and Engineering Applications

For the specified values of F and AX, the magnitude of AE is obtained by calculating N'from
Equation 9.39 and substituting the result into Equation 9.40. Both the X and E forces are inter-
nal and the sum of theseforces must satisfy the following two conditions:

\177]AX =0, \177AE=0 (9.4i)

The generalexpressionfor the factor of safety is obtained by adding the difference in the in-
terslice shear forces AX to the vertical forces in Equation 9.9:

1
F- . l
.i\177:Ic'b+[w(l-ru)+Zk\177]tanC#\" (9.42)
i=11. tact
y\177(wsin \177)i

The solution can be obtained by taking a set of X forces and a value of F that satisfy Equations
9.41 and 9.42. This requires increased iteration (in comparison with the simplified method)
and computational effort and must be carried out by a computer. Different assumptions for the
lines of action of the E forces may improve the factor of safety at the range of 1%.

9.6.3 Janbu's method

Janbu (1968 and 1973) proposeda general slope stability analysis that permits any shape of
failure surface to beconsidered. The analysis is based on the slice method and all equilibrium
conditions are fully satisfied. It also allows for variable lines of action for the interslice forces.
Similar to Bishop's simplified method, a minimum factor of safety is obtained for a critical
non-circular failure surface, but Janbu's analysis requires comparatively more computational
effort. While the search for a critical circuIar failure surface is relatively straightforward, the
search for a critical non-circular faiIure surface is considerably more complex and guidance
may be obtained from Celestino & Duncan (1981) and Chert & Shao (1988).Janbu's general

0 0.1 0.2 0.3 0.4

d/L

Figure 9.16. Janbu's method, (a) non-circular failure surface, (b) correction to the minimum factor of
safety.)))
 Stability of Earth Slopes 395

equation for the factor of safety is expressed by:

r
1
F - .c'b+(w-ub+ZLg)tan 4)'.
(9.43)
i
i=n(wtan\302\260t)i\177,
i=l
where all parameters have definitions identical to Bishop's method and maq) is defined by:

= cos 2 \177(i + tan \177.tan

\177b'/F) (9.44)
mc\177(j )

A method suitable for routine applications is based on neglecting the AX terms and increasing
the number of slices. To simplify calculations, the sfiding mass can be divided into several
verticat blocks, as Figure shown which are in turn further subdivided into finer
in 9.16(a),
slices. Consequently all the stices within a block will have a common ct value but different
pore pressures. Application of the routine method will cause a reduction in the factor of safety
corresponding to the criticat failure surface. Janbu's proposalto increase the factor of safety
for c' = 0, \177'and c', 0' = 0 soitsis shown in Figure 9.16(b). The correction factorfovaries with

the ratio of dWL, where L is the length AE shown in Figure 9.16(a) and d is the maximum dis-
tance of the non-circular failure surface from this line. The increased factor of safety is ex-

pressed by:

Fcor. = fo F (9.45)

For c', 9' soils,the average value offo corresponding to c' = 0 and = 0
\177\" curves can be used.
Exa\177nple 9.13
3
For a slope of H = 5 m, [3 c' =
= 45 \302\260, 15 kPa, 20.7 kN/m
\177b'
= 20
Ysat
\302\260,
= and Yd\177
= 17.5

kNhn 3, apply Janbu's routine \177nethod to the assu\177ned failure surface and with the piezo\177netric
level shown in Figure 9.17. Number of slices = 14,with b = 1 m. For the first trial assume the
factor of safety is 1.4.

Solution:
A summary of the co\177nputations for two trial values of the factor of safety is given in the tabte.
The finat factor of safety for the given non-circular failure surface (without using Equation
9.45) is 1.85.

C' D

/'Piezometric I F \177 xs= 4.0rn xc = 9.0 rn xD= | 4.0 m

xF=9.0 mxG= l&0 m

5.0
xE= 4.0 m
Failure YE
= 1.8 m
YF-- 3.2 m YG
= 4.0 m
\177, I/\177 I L,/
surface
C

Figure 9.17. Example 9.13.)))

396 Soil Mechanics: Basic Concepts and Engineering Applications

o: h h\177 w wtano: ub (w - ub)tant\177\" c'b

Slice tanc\177
(deg.) (m) (m) (kN) (kN) (kN) (kN) (kN)

i -I4.0 -0.25 0.625 0.35 12.06 -3.01 3.43 3.14 I5.0

2 -I4.0 -0.25 t.875 1.05 36.17 -9.04 t0.30 9.41 15.0
3 -14.0 -0.25 3.125 i.75 60.29 -!,5.07 17.[7 [5.69 15.0
4 -14.0 -0.25 4.375 2.45 84.40 -21.10 24.03 21.97 15.0
5 16.7 0\17730 5.35 2.79 102.55 30.77 27.37 27,36 15.0
6 I6.7 0.30 5.55 2.77 105.99 31.80 27.17 28.69 I5.0
7 I6,7 0.30 5.25 2.75 I00.67 30.20 26.98 26.82 I5.0
8 I6.7 0.30 4.95 2.73 95.36 28,61 26.78 24.96 15.0
9 16.7 0.30 4.65 2.7I 90.05 27.0I 26.58 23.10 I5.0
tO 42.0 0.90 4.05 2.35 78,39 70,55 23,05 20. I4 15.0
11 42.0 0.90 3.I5 1.65 60.40 54.36 I6.I9 16.09 15.0
12 42.0 0,90 2.25 0,95 42.41 38.17 9.32 I2.04 15.0
13 42+0 0.90 1.35 0.25 24.42 21.98 2.45 7.99 15.0
14 42.0 0.90 0.45 0,00 7.87 7.08 0.00 2.86 15.0
Total: 292.3t

F=1.4 F=I.85

tnc\177(t) [c'b+ (w - ub)tan\177'] / rectO') m\1779 [c'b+ (w - ub)tan\177'] / rnaq)

0,880 20,6I 0.895 20.27

0.880 27.74 0.895 27.27
0.880 34.87 0.895 34.29
0.880 42.01 0,895 41.31
0.989 42.83 0.972 43,58
0.989 44.17 0.972 44,95
0.989 42,28 0.972 43.02
0.989 40.40 0.972 41.11
0.989 38.52 0,972 39.20
0.682 5 t .52 0.650 54.06
0.682 45.59 0.650 47,83
0.682 39,65 0.650 41.60
0.682 33.71 0.650 35.37
0.682 26.19 0,650 27.48
Total = 530.09 Total = 541.34
F= 530.09/292.31 = 1,81 F= 541.34/292.31 = !.85

9.6.4 Morgenstern and Price's method

In this method the sliding above the non-circular failure

mass surface is first divided into wide
vertical slices (Figure 9.18(a)). Equilibrium of a vertical element of width d.x within each slice
(Figure 9.18(b))is then investigated following five functions:
by introducing
1. y = s(x) defines the st\177ape slope and of
its the
boundaries. For each slice this function is
given as data and may be represented by a linear equation.
2. y = y(x) defines the selected failure surface. For each slice a linear equation is assumed.)))
 Stability of Earth Slopes 397

line E
Pw\"\" F /\"\"t I
Vertical
I x I/X + dX
element / I il + dE'

\177 \177
..........
line
/ \177 \"\177
k \177dN

\177
/ ) X
Failure dP5 \177

Fibre 9.18. Morgenstem and P\177ce's method: (a) slo\177 s\177tion \177d wide slices, (b) fomes acting
vegicN element within a wide slice.

3. y = h(x) defines the line of action of the forces due to water pressure Pw acting on the verti
cal planes of each slice. This function is constructed from the position of the piezometric \177evel.
4. y = y't(x) defines the line of action of the effective normal force U acting on the vertical

planes.
5. X = )\177x)U defines the relationship between the shear force X and the effective normal
force U on the vertical planes of each slice. This makes the slope stability problem statically
determinate. The function f(x) is a given relationship that represents the ratio of the shear force
to the effective normal force on each vertical plane. X is a scale factor and is determined
within the solution. Force equilibrium equations, constructed parallel with and normal to the
sliding direction, are combined with an application of the Mohr-Coulomb failure criterion on
the failure plane to yield the following differential equation:

dP\177 dX dW dPb
C'sec2\177 ) =
-- 4 \"1-secO[
tan\302\242'(tanadE'+tane\177
F F c\177: \177 \177L\177 d\177 \177

dE\" dX dW
+ \177-
dP\177
tan \177z-- + tan a-- (9.46)

For convenience moment equilibrium is considered at the mid point of the base of the element
and yields the following differential equation:
\"
dE\" dP.,

Difficulties in the integration of the pair of the differential equations above may be overcome
by assuming linear equations for the five basic functions within the integration limits (wide
slices). This facilitates the integration of Equation 9.46 for selected values of F and \177.and

gives the E' forces as a function of x. The solution must satisfy the specified boundary condi-
tions in terms of the E' forcesat the first and last slices. Using the same basic five linear func-)))
398 Soil Mechanics: Basic Concepts and Engineering Applications

tions, 9.47 is integrated

Equation to yield the change in the moment, which is essentially zero
or as specified first and last at the
slices. If the solution does not satisfy the moment equilib-
rium conditions another set of values for F and Z must be tried. Estimates of these new values
can be obtained using a two variable Newton approximation method that gives the increments
AF and A2 to be added to the previous values of F and \177.

9.6.5 Spencer's method

This analysis is based on the sIice method and is applicable to failure surfaces of any shape
and satisfies all conditions of equilibrium. Its application to circular failure surfaces has shed
some light on expIaining the accuracy of Bishop's simplified method. Referring to Figure
9.5(b), the inter-slice forces XI, X2, El and E2 can be represented by a single force Z with its
point of application on the centre line of the slice and making an angle of 0 with the horizon-
tal. For the general formulation it may be assumed that the force Z is a disturbing force in a di-
rection opposite to the force T. Vertical and horizontal force equilibrium equations are com-
bined (to eliminate N') and result in the following expression for Z:

c'l/F+ cosc\177
- W sin\177
(W -ut)tan\177'/F
Z = (9.48)
- -
cos(c\177 0)[1 + tan(c\177 0) tan 9\"! F]

From the definition force Z, the difference

of the in the inter-slice forces at sides of the slice is
expressed by the following two equations:

\177 = Zsin0, AE = Zoos0 (9.49)

Substituting the expressions above into Equations 9.41 we obtain:

0 = 0, y\177Zcos0
-- 0 (9.50)
\177Zsin

Moment equilibrium of the Z forces about the centre of rotation of the slice gives:

Z cos(c\177 - O) =0 (9.51)
\177

An iterative method is used to solve Equations 9.48 and 9.50. An equal 0 value for all slices

implies that the resultants of the side forces are parallel.Equation 9.50 then reduces to:

z = 0
\177 (9.s2)
In the case of a constant 0, the solution may be obtained as follows:
1. Assume a value for the unknown angle 0 and find the Z force for each sliceusing Equation
9.48.
2. Substitute the Z values into Equation 9.51 and evaluate F using a trial and error method.
3. If the assumed value of 0 and the computed value of F satisfy Equation 9.52, the solution is
correct; otherwisea new value for 0 must be selected.
It may be shown that the factor of safety is not very sensitive to variations in 0 which explains
the accuracy of Bishop's simplified method. For 0 = 0 the factor of safety is identicalto that
predicted by Bishop's method.)))
 Stability of Earth Slopes 399

9.7 APPLICATION OF THE WEDGE SLOPES

METHOD TO UNREINFORCED

9.7.1 Hard stratum in the vicbffty of an earth slope

Single or multiple wedge are applied to unreinforced

mechanisms sIopes where there is the
possibility of translational sliding on the surface of a hard stratum close to the ground sud'ace.
Possible sliding mechanisms of an earth slope in the presence of a hard stratum are shown in
Figure 9.19. In Figures 9.19(a), (b) and (c), the surface of the hard stratum is incIined in the di-
rection of the slope and may intersect the sloping ground, or pass through the toe or be located
at some depth below the toe. Whilst the first two cases may be analysed using a single wedge,
a multiple wedge mechanism is more appropriate for the third case. Figures 9.19(d) to 9.!.9(f)
show the mechanisms for the particular instance of a horizontal bard stratum. The first two
cases can be anaIysed using a single wedge but the accuracy of the result depends on the slope
angle, The third case is a
analysed using multiple wedge mechanism or a circular method.

x AC
.... T
/-/\177 \\ \177H /.\177'/\177 Hard
Hard N' \177 _\177'/__\177\"/. stratum
stratum \">\"\177\">'\177
A

(a) (b)

C\177
B

\177-surface
\177H \177
\177m

stra,um

\177'\" \177

Hard stratum Har\177 stratum

(e) (f)

Figure 9.19. Application of the wedge method to unreinforced slopes in the presence of a hard stratum.)))
400 Soil Mechanics: Basic Concepts and Engineering Applications

For a single wedge the stability of the slope can be formulated in terms of the stability number

\177H / height or criticaI slope al\177gle. In the case of a multiple

c', critical wedge mechanism the
factor of safety is defined as the ratio of the shear strength on the sliding surface to the shear
stress on that surface required to maintain equilibrium. In a c' = 0, \177' soil this ratio becomes:

F = zf/'c =N tan\177/N tannin

=
tan\177/tan%\177 (9.53)

where 9\"m is the mobilized friction angle on the interface of the soil and the hard stratum that
satisfies the force equilibrium equations. It is assumed that the maximum friction angle that

can be mobilized sliding on the is equal to the internaI friction angle of the slope mate-
surface
rial. In order to simplify calculations and ensure a conservative factor of safety, the friction

angle on the vertical inter-wedge surfaces may be assumed to be zero. For a given mechanism,
the geometry of the wedges must be optimised to obtain the minimum factor of safety.

9.7.2 Singlewedge mechanisrn

By assuming full mobilization of the shear strength parameters for a specified slope angle [3,
the stability number corresponding to the critical height can be evaluated.Referring to Figure

9A9(a) and considering force equilibrium in the AC and BD (normal to AC) directions we ob-
tain:

N'tanO'+c'\303\227AC-Wsinct=O, N'-Wcosc\177 =0 (9.54)

Combining the two above equations and replacing Wwith H c sin(t3-ct)ACy/(2sinl3) gives:
yH,. 2sin[3cosqb'
(9.55)
c\" -
(x) sin(ct
-
sin(]\177 \177')

It is seen that the stability number is independent of the slope of the upper ground surface. If a
factor of safety is specified for the shear strength parameters, then:

yH 2sin \1773cosO\177,
(9.56)
- -
% sin([3 ct) sm(ct %, )
where C'm = mobilized effective cohesion =c' ! F, and 9',, is the mobilized friction angle on the
sliding surface so that tan\177'm
= (tanqb') ! F. For the case where H < He, a factor of safety can be
derived by replacing N' tan \177'+ c'x AC with (N' tan \177'+ c'x AC)! F in Equation 9.54 and
rearranging the result thus:

F - 2c'sinl3 tan0'
\177
- \177- (9.57)
yH sin(13 a) sin c\177 tan c\177

In general the analysis of a singlewedgeis mathematically identical to an upper bound analy-

sis where force equilibrium and energy equilibrium both yield the same result. In a c\" =0, 9'
soil, the factor of safety is expressed by the second term of Equation 9.57, which is identical to
Equation 9.19 and is independent of the height and the slope angle. In this case, the value of F
> 1 means the slope will not slide on the hard stratum but it may slide on anotl\177er plane or fail-)))
 Stability of Earth Slopes 401

ure surface within the cohesionless material. In undrained conditions Equation 9.57 is rear-
ranged (with F = 1) to the following form:

2 sin
= \177'H\177= 13
(9.58)
Nf cu sin([3- \177)sin \177x

In this case Equation 9.5 can be usedto evaluate the factor of safety.

9,7.3 Multiple wedge mechanism

Multiple wedge mechanisms are used for the cases shown in Figures 9.19(c) and 9.19(f) by
considering the general definition of the factor of safety for a limit equilibrium method and
maximum values of the mobilized shear strength parameters, To begin with, an iterative
method is used to verify the assumed factor of safety for the mechanism, The next step is to
optimise the variables of the mechanism to obtain the minimum factor of safety. On the inter-
wedge surfaces the mobilized cohesion and friction angle must also be specified.A common
approach to the solution is to guess the factor of safety and evaluate the force equilibrium of
each wedge starting from one side of the mechanism and moving towards the last wedge on
the other side. The interface forces calculatedfor the last wedge must satisfy equilibrium of
this wedge. After the first iteration the residual interface forces for the last wedge can be used
to formulate the variation of the mobilized strength parameters in terms of the computed reac-
tions on the sliding surfaces. The mobilized strength parameters should be changed in a sys-
tematic way until lbrce equilibrium of the last wedge is satisfied,

Example 9.14
A with = 45 \302\260
be constructed on the surface
is to of a hard stratum that has an angle of
slope 13
30 \302\260
to the The soil propertiesare: c' = 10kPa, 0' = 20 \302\260
horizontal, and y = 18 kN/m 3. Assum-

ing that the soil abovethe surface of the hard stratum slides on this surface, determine: (a) the
critical height and (b) the corresponding height for a factor of safety of 2.

Solution:
Use Equations 9.55 for case (a) and 9,56 for case (b):

(a) Nf - Yffc - =
180H-\177-.-\302\243-c
2\303\227sin45'0\302\260\303\227c\302\260s20'0\302\260-
29.57 --\177H = 16.4 m.
c
c' 10.0 sin(45.0 \302\260
- 30.0 sin(30.0
\302\260)
\302\260
- 20.0 \302\260)

(b) cm = c'/F = 10.0/2 = 5.0kPa, tan %, = tan 9'/F = tan 20.0\"/2 -\177
Cm
= 10-3\302\260,

- yH _ 18-0Hc 2\303\227sin45.0\302\260\303\227cost0.3
\302\260
-
N = 15.95 \177 H = 4.4 m.
\302\260
- \302\260- c
c\177,
5.0 sin(45.0 30.0 sin(30,0
\302\260) 10.3 \302\260)

Example 9.15
\302\260
= 45 \302\260 stratum inclined
Figure 9.20(a) shows a slope with 13 constructed on a hard at \177= 31t to

the horizontal. Calculate the factor of safety for the selected three-wedge mechanism. Lines)))
402 Soil Mechanics: Basic Concepts and Eagineerbzg Applications

FB and EC are verticaI. The soil properties are: c' = 0, d)'

= 36 \302\260
and 1' = 18 kN/m 3. Assume

there is no frictional forces on the vertical interfaces of the wedges.

SoIution:
2
The areas of the wedges are calculated from their geometry: A1 = 4.076 m 2, A2
= 15.0 m
A3
= 3.925 m . Thus the corresponding weights are:
W r = 4.076 xl.0x 18.0 =-73.37 kN, W
2
= I5.0xl.0xlS.0 = 270,0 kN,
W3 = 3.925\303\2271.0\303\227i8.0 = 70.65 kN. A summary of the computations is tabulated below, where

for = 24.5 \302\260,

\177',\177 E2R = EeL and the factor of safety for the assumed mechanism is 1.59.

O'm (deg.) Et (kN) E:\177 (kN) Ezt(kN) F

36.0 37.38 ]3.76 138,66 1.00

25.0 57.32 85.70 90.43 !.56
24.5 58.36 89.12 88.82 1.59

\302\260
Sample calculation for d\177'rn= 25.0\302\260:F = tan 0'/tan \177 = tan 36,0\302\260/tan 25.0 = 1.56.
Force equilibrium of wedge1 in the directions CD and normal to CD gives:
-
\302\260 \302\260 \302\260 \302\260
- \302\260
N\177
tan 25,0 W1 sin 63.0 + E\177cos 63,0 = 0, N\177 - W] cos 63.0 Er sin 63.0 = 0.
Solving for E\177:

E 1
W\177(tan63'0\302\260-tan25\"0\302\260)
= 73\"37(tan63\"0\302\260-tan25\"0\302\260)
\302\260 =57.32kN, similarly:
1 + tan tan
63.0\302\260 25.0
\302\260
1+ tan 63.0 \302\260
tan 25.0

31.0 -
\302\260
tan 25.0 \302\260)
E2R
= 270.0(tan
\177
\177-
Er
= 28.38 + 57.32 = 85.70\177.
I+ tan 31.0\177tan 25.0

In the third wedge force equilibrium yieIds:

\302\260
W 3 (tan 27.0 + \177n 25.0 \177) 70.65(tan 27.0
\302\260
+ tan 25.0 \302\260)
= = 90.43 \177 > 85.70 \177.
E2L= o
l _ tan
\177
27.0O tan 25 .0 l_tan27.0Otan25.0

Thus the mobilized friction angIe on the sIiding surfaces must be slightly decreased to main-
\177in force equilibrium.

\177 ( 2 /\177 , ,
(---\177 Hard
\177 E2c E2R
2.0 2tane m
\177\177t;a\177um \177
\177\177/\177N
\177
(at \177N,
a
(b)

Figure 9.20. Example 9.15.)))

 Stability of Earth Slopes 403

9.8 CONCLUDING REMARKS

9.8.1 Estimation of settlement in slopes and earth dams

In slope stability analyses that use limit equilibrium methods, the state of stresswithin the slid-

ing mass is unknown. Estimation of settlements in slopes and earth dams may be made using
methods based on continuum mechauics (e.g. ABAQUS finite element code by Hibbit, Karls-
son & Sorensen,Inc.,1995). Simplifying assumptions, such as a linear distribution of vertical

stress with depth, facilitates estimation of the deformation and settlement of the sliding mass.
For one-dimensional displacements (oedometertype), the distribution of vertical displacement
can be formulated by assuming a linear distribution of vertical stress. For the plane strain case

the semi-numerical method described in Chapter 5 (Equation 5.75) may be used. In this case

the slope is divided into horizontal layers and estimates are made of the vertical and lateral
stresses at the centre of each layer. Hooke's law can then be used to estimate of the vertical de-
formation of each layer.
A simplified one-dimensional analysis, applicable to the construction of an earth dam with
symmetrical slopes, is given by Marsal (i958) who suggested using a paraboIic shape for the
elastic vertical displacement under the centre line. At elevation z measured from the founda-
tion level, the vertical stress is y(H- z) where \302\245 is the unit weight of the slope material. Con-
sidering the simplified definition of strain, the vertical displacement v(z) is given by:

cr_ v(z)
y(H-z)
\177v(z)= y__\177z(H_z) (9.59)
Es E\177. Z

where E.\177is the Modulus of Elasticity of the slope material. From the parabolic function ex-
pressed by Equation 9.59, it can be seen that the vertical displacement is maximum at mid-

height of the dam and is zero value at the upper ground surface and the foundation level. Sub-
stituting H / 2 for z we obtain:
2
Vma x
=
yH / 4E s (9.60)
Although the conditions assumed in this model are unrealistic,
this simpIified analysis can be
used to estimate vertical displacements in the stages of construction
early and assist in the in-
terpretation of results obtained from the plane strain models (Naylor, 1991). A modified
method assumes a linear variation for E\177.and y with depth (Pagano et al., 1998). For a constant

y, a linear expression for Es may be defined by: E s = E\177 +\177qz, where Eo is the Modulus of

Elasticity at the foundation level. The vertical displacement at elevation z is calculated from:
= -
v(z) [y(H z) lrl] In(1 + \177]z! E
o) (9.61)
For constant Es, a linear expression for y is defined by: y
=
Yo -?\177z, where yo is the unit

weight at the foundation level. The vertical displacement at elevation z is then given by:

v(z) = \1771 [2yo(H

- z)z \177(H
2 _
z2)z] (9.62)
2E s)))
404 Soil Mechanics: Basic Concepts and Engineering Applications

For positive values of rl and )\177,the maximum vertical displacement occurs belowmid-height
whilst for negative values it occurs above mid-height.

9.8.2 Application of numerical methods to slope stability

With the advent of the new generation of powerful computers, the finite element and explicit
finite difference methods have been applied successfully to a variety of slope stability prob-
lems. A comprehensive history of the finite element analysis of slope stability is presented by
Duncan (1996) and the following conclusions were drawn:
1. An elastic stress-strain model (linear, muttilinear, and hyperbolic) represents soil behaviour
well at low levels of stressbut is unable to predict the plastic deformation. Difficulties in the
determination of Poisson's ratio and the Modulus of Elasticity are the major problems,
2. The totat deformation (elastic plus plastic) can be predictedwith reasonable accuracy using
elastic-plastic models or non-Iinear stress-strainmodels.
3.The finite element method overestimates the deformation because soit properties from labo-
ratory tests are used. Ageing of the soil in the field and disturbance of the sample in the test
are the major reasonsfor this outcome.
4. The cost of using finite element method is about 10% of the total cost of an analysis.
The stability ofa verticaI cut for ru= 0 is shown in Tables 7.1(a) and 7.1(b), (Chapter 7). It
is seen that, for \177)'values less than 40 \302\260,there is little difference between the slice solution and
the two numerical techniques of finite element method (FEM: AFENA package by Carter &
Balaam, 1990)and explicit finite difference method (FDM: FLAC package version 3.2, 1993
by Cur\177dall, 1980 & 1987)- Results obtair\177ed from the mechanism and slice methods are almost
identical. All results fatt within the upper bound and the lower bound solutions obtained by fi-
nite element formulations (Aysen & Sloar\177, 1992). An elastic-plastic stress strain model has
been used in both FEM and FDM applications to study the stability of slopes. For slopeswith
small angles, the FEMJFDM results always exceed those of the slice solution (higher stability
\302\260
number or toad parameter). Typically these differences are negligible for \177' < 20 but signifi-
cant (say 25% or more) for \177' > 20 \302\260. But this generalization must be viewed with caution as
the results from numerical methods depend upon the selected modeI, its corresponding pa-
rameters and the technique in which the ultimate failure load is determined in the incrementaI

process of the computations (Carter & Balaam, i990).

In the case where yH / c' = 0 (a weightless material), a uniformly distributed load q applied
to the upper ground surface is increased incrementally using a dunwny time variable which
calculates the proportion of the specified loads acting on the top of the slope at each computa-
tion step. The accuracyof the method depends on the number of increments and interpretation
of the stress-strain (or load-deformation) behaviour of some selected elements. Usually the

failure load for this case corresponds to an etemem located at the intersection of the upper
ground surface and the inclined face of the slope. A similar procedure must be carried out for
some other elements that are thought to be in a state of faiturer In the case of self-weight with
a zero surface load, techniques are adopted to simulate excavatior\177 arid fill processes. For ex-
ample, to simulate the fill slope case, a progressive increasein yH / c\" may be achieved by
adding to the flat layer a set of 6-8 rows of elemer\177ts, with each set fbnning a uniform thick-)))
 Stability of Earth Slopes 405

hessof soil layer to be added as an increment of fill (Aysen & Loadwick, 1995).Displacement
behaviour of the toe element may be used to calculate the final value for the stability number
at failure.
In the numerical application or FDM, the initial state of stress within
of FEM the domain of
the problem is required input data. The vertical stress can be assumed
as to be equal to the

overburden pressure while the horizontal normal stress is estimated using the concept of the
at-rest condition (Chapter 8). If required the initial strains are set to zero.

9.8.3 Three-dimensional slope stability analysis

Whilst the slope stability analyses in this chapter are based on the plane strain conditions,
many practical cases should be treated as a three-dimensional stability problem. Such analyses
include limit equilibrium and variational methods (assuming a polynomial equation for the
failure surface) such as extended slice methods (Fellenius, Bishop's simplified method, Janbu
and Spencer), extended Morgenstern and Price method, upper bound solutions and numerical
methods. Reference may be made to Hovland (1977: extended Feltenius), Huuger (1987: ex-
tended Bishop's simplified method), Chen & Chameau (1982: extended Spencermethod and
finite elements), Chert & Chameau (1983:extended Spencer method), Leshchinsky & Huang
(1992: limit equilibrium and variational method). Generally the three dimensional factor of
safety F3 is greater than the two dimensional factor of safety F2, where the difference depends
on the shear strength parameters and the length of the extension of the failure surface from the

plane strain into the third dimension, The ratio F3 / F,\177in a c\" = 0, \177' soil is normally in the
range of 1.0 to 1.08 but tnay rise to 1.30 as the cohesion increases (Azzoz et at., 1981).In
some cases the ratio has been found to be less than 1.0 (Chert & Chameau, 1983;Seedet at.,

1990). This is believed to be a result of simplifying assumptions made in special cases but this
has yet to be proven.

9.8.4 Generalobservations
The results obtained by circular methods for a c,, \177' = 0 material may differ from those of
more sophisticatedanalyses by some 5% to 8% (Bromhead, 1992).Duncan (1996) reported a
maximum variation of 12%between the results obtained for rotational sliding using the differ-
ent approaches presented in this chapter. Generally it seems reasonable to use Bishop's simpli-
fied method with a large number of slices (say 30) and start the iterative process with a factor
of safety slightly higher than the one obtained (for the failure surface) from Fet-
satne circular
tenius' method. Settlement of the slope under its own weight and any boundary loading can be
evaluated using the method of Section 9.8.1. Alternatively, a numerical lnethod with an appro-
priate stress-strain tnodel obtained in the laboratory condition may be adopted. If settlement is
not an important factor and it is required to evaluate the stability using FEM or FDM, an elas-
tic-plastic stress-strain model can be used. In cut slopes the heave of the lower ground surface,
which occurs due to unloading, must be modelted using a stress-strain relationship that in-
cludes both toading and unloading cycles. Whilst an estimate of the heave is possibleusing the

elastic methods presented in Chapter 5, the use of FEM or FDM anatyses is preferred. Con-)))
406 Soil Mechanics: Basic Concepts and Engineering Applications

solidation of the foundation soil may be estimated using the principles of Chapter 6, A coupled
solution, where the stability of slope and consolidation of the foundation soil are analysed un-
der a time-related loading, must be carried out by numerical methods. A crude estinaate, how-

ever, can be obtained by applying the principles of Chapter 6 with time related loading. It is
preferable to obtain pore pressures from the flow net, otherwise an average pore pressuremust
be evaluated.

9.9 PROBLEMS

9.1 Determinethe factor of safety for a 1 vertical to 2 horizontal slope 5 m high using a trial
toe circle for which
:\177
xc = 4.5 m and yc = 8 m. The cross-sectional area of the sliding
mass is 40,22 m and its centroid is located 2.69 m from the centre of the trial circle. The
soil properties are: cu = 18 kPa, \177u
= 0, and =
\177' 18 kN/m 3.
Answer: 1.36
9.2 For a 45
\302\260
plane strain slope 30 m high (Example 7.8), determine: (a) the factor of safety
for a toe circlefor which xc = 12.5 m and YC = 42 m if there is no surcharge load on the
upper ground surface, and (b) find the maximum surcharge load q that will cause the
failure of the or\177the same slip circle. The soil properties are: c, = 100 kPa, dou = 0,
slope
and y = 18 kN/m .
Answers: 1.09, 45 kPa.
9.3 Using Taylor's stability chart (Figure 9.4), re-solvepart (a) of Problem 9.2.
Answer: 1.02
9.4 Using Fellenius' determine the factor of safety
method for a slope of 1 vertical to 2 hori-
zontal and height m using a trial toe circle for which xc -4.5 m and yc = 6.25
H =4.5
m. The soil \177nass is divided into 3 m wide slices whoseaverage height and angle \177 are
tabulated below, c' -- 6.75 kPa, d0\" = 17 \302\260,
and 9 = 1.96 Mg/m3.

Sliceno. 1 2 3 4
h (m) !.6 3.7 4.6 3.0
\177(deg.) -23 0 23 51
Answer: 1.49
9.5 A 5 m high slope has an angle of [\177= 45 \302\260.
Data on the 1 m wide slices are given in the

table below where hw is the height of water measured from the midpoint of the base of
each slice. The trial circle is not a toe circle and slices I and 2 are \177ocated to the left of
the toe. Using Fellenius' method, determine the factor of safety for this trial circle, c'
= 15.0 kPa, \177\"= 20 \302\260,
\177'sar
= 20.7 kN/m 3, and Yd
= 17.5 kN/m 3.
Answer: 1.465

9.6 An 8 m slope for which = 35 \302\260

is composed of two of soil with the
13 layers following

properties:)))
 Stability of Earth Slopes 407

Sliceno. h (m) h\177,,(m) o\177

(deg.)

1 0,20 0.20 -24.0

2 0.60 0,60 -14.0
3 1.35 1.35 -11.0
4 2.40 2.40 -3.0
5 3.40 3.20 0.0
6 4.35 3.60 5.5
7 5.25 3.80 11.5
8 5.60 3.80 14.0
9 5.25 3.70 24+0
10 4.75 3.40 29.0
11 4.20 3.10 32,5
12 3.50 2.60 38.5
13 2.50 !.70 46.0
14 1,25 0.60 57.0

5m = 25,0
kPa, do'] = 12\302\260,\1771 = 18 kN/m 3.
Upper layer: thick, c']
Lower layer: 3 m thick, c'2 =
7.0 kPa, do'2 = 25 \302\260, Y2
= 19.5 kN/m 3.
The centre of the trial circle is located 5 m to the right and 12 m above the toe and its ra-
dius is 13.89 m. Consequently the failure surface is not a toe circle but passes from a
point 2 m to the left of the toe. The total central
\302\260
angle is 103.6 for which the central an-
\302\260
gles correspondingto the upper and lower layers are 23.7 and 79.9 respectively.
\302\260
Data
on the slices are tabulated below. Slices I to 9 are 2 m wide whilst the width of slice 10
is 2.4m. Calculate the weight of each slice from the equation w = b(ylhl + y2h2), where

hi, h2 are the heights corresponding to layers 1 and 2 respectively. The angle c\177for each
shce is to be computed from a =sm (x / R), where x Is the hortzontal distance from the
midpoi,lt of the base to the centre of the circle. Using Fe[lenius' method, calculate the

factor of safety for the given trial circle.

Slice no. 1 2 3 4 5 6 7 8 9 10

tq (m) 0+00 0.00 0.00 0.55 2.00 3.30 4.70 5.00 5.00 2.90
h2 (m) 0.60 2.10 3.90 5.00 4.80 4.40 3.50 2.50 0.70 0.00

Answer: 1.25

9.7 A slope of I vertical to 2 horizontal and height of 7.5 m has the following soil properties:
c' = 15 kPa, d0' = 25 \302\260
and y = 20 kN/m 3. Using the stability coefficients of Bishop &
Morgensten\177 (1960), compute the factor of safety for ru = 0.0, 0.2, and 0.4.

Answers: 2.19, 1.86, 1.54

9.8 A
ton\177g
slope is to be constructed using a material with: c' =0, q\177'= 35 \302\260,
and Y.,at
= 20
kN/mZ Determine the critical angles for both the dry condition and steady state flow
parallel to the surface. Calculate the factor of safety for both cases if the selected slope
angles are equal to f5,,/1.5

Answers: 35.0 \302\260,

19.6 1.62,
\302\260, 1.54)))
408 Soil Mechanics: Basic Concepts and Engineering Applications

13 C ED

\"
.0 m
\177-
\"\1774

(a) 2.0roT [ t---

Figure 9.21- (a) Problem 9.10, (b) Problem 9.11.

9.9 Re-work Example 9.10 using a toe circle of radius 10.54 m tangent to the base at the toe.
Relevant data are given in the table below.

Sliceno. I 2 3 4 5

h (m) 1.7 4.9 5.9 4.4 1.8

ct(deg.) 5.4 16.6 28.4 41.8 59.5

Answer: T\177otal
= 250.8 k.N

9.i0 For thereinforced slope shown in Figure 9.21(a), calculate the total force in the rein-
forcement for the a trial two-part wedge shown. The soil propertiesare: c' = 0, 0' = 29\302\260,
and y = 18 kN/m 3.

Answer: 59.5 kN
9.11 For the multiple-wedge mechanism shown in Figure 9.21(b), calculate the factor of
safety for the slope assuming that no cohesion or friction is mobilized on the vertical in-
ter-wedge planes of CE and BF. c' = 9.5 kPa, 9' = 30.8\302\260,and = 18
\177' kN/m 3.
Answer: 1.9

9.10 REFERENCES

Atkinson, 3.H. 1993.An introduction to tlae mechanics of soils and foundations. London: McGraw-Hill.
tj'
Aysen, A. & Sloan, S,W. 1992. Stability of slopes in cohesive frictional soil. Proc. 6 Australia-New
Zealand co\177 on geomechanics: Geotechnical risk-identification, evaluation and solutions: 414-419.
New Zealand: New Zealand Geomechanics Society.
Aysen, A. & Loadwick F. of slopes in cohesive frictional
1995.
S\177ability soil using upper bound collapse
mechanisms and methods, Proc. 14 Australasian
numerical conf on the rnechantcs of structures a,\177d

materials 1 : 55-59. Hobart, Australia: University of Hobart

Azzoz, A.S., Baligh, M.M. & Ladd, C.C. 1981.Three dimensional stability analysis of four embankment
\342\200\242 th .
fadures. Proc. 10 intern, cot\177 SMFE 3: 343-346. Rotterdam: A.A. Balkema.
Bishop, A.W. 1955.The use of slip circle in the stability analysis of slopes. Geotechnique 5(1):7-17.
Bishop, A.W. & Morgenstem, N.R. 1960.Stability coefficients for earth slopes. Geotechnique 10(4):
129-147,)))
 Stability of Earth Slopes 409

\342\200\242
. nd . .
Bromhead, E.N. 1992. The stability of slopes. 2 edition, Surrey: Surrey University Press.
Carter, J.P., & Balaam, N.P. 1990. Program AFENA: A generatfinite element algorithm. University of

Sydney, Australia: Centre for Geotechnical Research.

\177
Celestino, T.B. & Duncan, J.M. 1981. Simplified search for non-circular slip surface. Proc. t\177 intern.

cor\177 on soil mech. and found. Engineerb\177g 3: 391-394. Rotterdam: A. A. Balkema.

Chandler, R.J. & Peiris extensions to the Bishop
T.A.1989. Further and Morgenstern slope stability
charts. Grot\177nd
engineering, May: 74-91.
rt)
Chert, R.H. & Chameau, J.L. 1982. Three dimensional slope stability analysis. Proc. 4 intern, conf nu-
met. meth. b\177geomech. 2: 671-677. Rotterdam: A. A. Balkema.
Chen, R.H. & Chameau, J.L. 1983. Three dimensional limit equilibrium analysis of slopes. Geotechnique
33(1):31-40.
Chen, Z. & Shao, C. 1988.Evaluation of minimum factor of safety in slope stability analysis. Canadian
geotechnical journa125(4): 735-748.
Cousins, B.F. 1978. Stability charts for simple earth s]opes\177 Journal GE, ASCE 104.
Cundall, P.A. 1980. UDEC - A generalized distinct element program for modelling jointed rock. Peter
Cundall Associates, Report PCAR-1-80, US Army, European research office.
Cundzll, P.A. 1987. Distinct element models of rock and soil structure. In E. T. Brown (ed), Analytical
and computational methods in engineering rock mechanics: 129-139. London: George Allen & Un-
win.

Duncan, J.M. 1996. State of the art: Limit equilibrium and finite element analysis of slopes. Journal GE,
ASCE 122(7): 577-596.
Duncan, J.M., Buchignani, A.L. & De Wet, M. 1987.An engineering manual for slope stability studies.

Blacksburg, VA: Virginia Tech.

nd
Exxon Chemicals. 1992. Geotextiles: design for soil reinforcement. 2 edition: 58. UK: Exxon Chemi-
calsGeopolymers Ltd.
W. 1927. Erdstatische
Fel[enius, berechnungen mit reibung und kohasion. Berlin: Ernst, (in German).
Hibbit, Karlsson & Sorensen, Inc. 1995. ABAQUS/ Standard - user manual - version 5.5. Pawtucket,
R.I.: Hibbit, Karlsson lnc.
& Sorensen,
Hovland, H.J. 1977. Three dimensional slope stability \177nalysis method. Journal GE, ASCE 103(9):971-
986. ASCE.

Hunger, O. 1987. An extension of Bishop's simplified method of slope stability analysis to three dimen-
sions. Geotechnique 37(1): 113-117.
lngold, T.S. 1994.The geotextiles and geomembranes manual. Oxford, UK: Elsevier Advanced Tech-
nology.
Janbu, N. 1968. Slope s\177ability Soil mechanics and foundation engineering
computations. report. Trond-
helm: Norway, The University of Norway.
Technical
Janbu, N. 1973. Slope stability computations. In E. Hirschfield & S. Poulos (eds), Embankment dam en-
gbteering, Casagrande memorial \177olume: 47-86. New York: John Wiley.
Jewel], R.A. 1991. Application of revised design charts for steep reinforced slopes. Geotextiles & ge-
ornembranes, 10(3): 203-233. UK: Elsevier.
Jewe]l, R.A. 1996. Soil re\177nforce\177ent with geotext\177les. Construction industry research and information
association (CIRIA). Special publication 123. UK: Thomas Telford.
.lewell,R.A., Paine, N. & Woods, R.I. 1985,Design methods for steep re\177nforced embankments, polymer
grid reinforcement: 70-8I. UK: Thomas Telford.
King. C.LW. 1989.Revision of effective stress method of slices.Geo\177echnique 39(3): 497-502,)))
410 Soil Mechanics: BasicConcepts and Engineering Applications

Koerner, R.M. 1984.Slope stabilization using anchored geotextiles: Anchored spider netting. Proc. spe-
cial geotechnicat engineering for roads and bridges conf : 1-11,Harrisburg, PA.: Penn DOT.
Koerner, R.M. 1991-Geomembranes overview, siglaificance and background. In A. Rollin & J.M. Rigo
(eds),Geomembranes identifications and performance testing. Chapter 1 : 355. London: Chapman &
Hall.
rd
Koerner, R.M- 1994. Designing with geosynthetics. 3 edition. London: Prentice Hall.
Koerner, R.M. & Robins, J.C. 1986.In-situ stabilization of soil slopes using nailed geosynthetics. Proc.
ru
3 conf on geosynthetics: 395-399. Vienna.
Ladd, C.C. 1991. Stability evaluation during staged construction. Journal GE, ASCE 117(4): 537-615.
Leshchinsky, D. & Huang, C. 1992.Generalized t\177ee dimensional slope stability analysis. Journal GE,
ASCE 118(1 !): 1748-1764. ASCE.
Marsal, R.J. 1958. Analisis de asentamientos en la presa Presidente Ateman. Oaxaca(5).htstituto de ln-

genieria. Mexico City, Mexico: UNAM.

Morgenstern, N.R. & Price, V.E. 1965. The analysis of the stability of general slip surfaces. Geotechni-
que 15(!):79-93.
Naylor, D..1. 1991. Finite element methods for fills and embankment dams. In M. das Neves (ed), Ad-
vances structures. North Atlantic Treaty Organization advanced study institute series: 291-
in rockfill
339. Dordrecht, The Netherlands: Kluver AcademicPublisher.
Pagano, L., Desideri, A. & Vinale, F. 1998. Interpreting settlement profiles of earth dams. Jo,\177rnaf GGE,
ASCE 124(10): 923-932.
Schaefer, V.R., Abramson, L.W., Drumheller, J.C., Hussin, J.D. & Sharp K.D. 1997.Ground improve-
ment, ground reinforcement, ground treatment-developments 1987-1997. ASCE Geotechnical special
publication, (69): 263-270. New York: ASCE-
Seed, R.B.,Mitchell, J.K. & Seed, H.B.1990. Kettlemen Hills waste landfill slope failure, 11: stability

analysis. Journal GED, ASCE. 116(4): 669-690.

Spencer, E. 1967. A method of analysis of the stability of embankments assuming parallel interslice
forces. Geotechnique 17(! ): ! 1-26.
Tavenas, F., Trak, B. & Leroueil, S. 1980.Remarks on the validity of stability analyses. Canadian geo-
technical journal, 17(I): 61-73.
Taylor, D.W. 1948.Fundamentals of soil mechanics. New York: Wiley.
UK Department of Transport. 1994. Design methods for the reinforcement of highway slopes by rein-

Whitlow,
\342\200\242 .
R. 1990. Bastc soil mechamcs.2 edttlon,
rid ....
forced soil and soil nailing techniques. Design manual 4, section 1, HA 68/94.
New York: Longman Scientific & Technical.)))
CHAPTER 10

Bearing Capacity of Shallow Foundations and Piles

10.1 INTRODUCTION

Shallow foundations are comprised of footings and rafts, while the deep foundations include

piles that are used when the soil near the ground surface has not adequate strength to stand the
applied loading. The ultinvate bearing capacity (in kPa) is the load that causes the shear failure
of the soil immediately underneath and adjacent to the footing. A relatively high factor of
safety (3 or more) is applied to allow for uncertainties in soil properties, inaccuracies in the

method of analysis and allowable settlement recommended by the building codes or the de-

signer. To calculate the ultimate bearing capacity for a geveral shear failure a failure mecha-
nism is introduced that satisfies equilibrium, stress boundary conditions and the failure crite-
rion. The failure criterion is satisfied on the sliding surface(s) as well as inside the individual
blocks of the mechanism creating a plastic equilibrium in the shearing zone. In practice this
failure occurs suddenly and is catastrophic. Sometimes the failure is associated by a significant
vertical settlement and failure planes do not reach to the ground surface. This is referred to as
a punchirrg shear failure because the soil around the footing is relatively uninvolved in the
failure. In a local shear failure a significant settlement irrtmediately beneath the footing is as-

sociated with the upward movement of soil at both sides of the footing and the failure planes

may not reach the ground surface. Thus the failure pattern has the characteristics of general
and punching shear representing a transitional mode (Ramiah & Chickanagappa, 982).It is 1
proposed that the ultimate bearing capacity for local shear to be computed based on 1 / 3 re-
duction of the shear strength parameters. In-situ tests such as plate bearing test, Standard Perpe-
tration Test (SPT),and Cone Penetration Test (CPT) are performed to estimate the allowable

bearing pressure and ultimate bearing capacity.

[0.2 ULTIMATE BEARING CAPACITY OF SHALLOW FOUNDATIONS

10.2.1 Uttimate bearing capacity offrictiontess shallow strt\177 footing due to Parndtl
A failure mechanism for a shallow strip footing proposed by Prandtl in early 1920s is shown
in Figure 10.1. The footing is assumed frictionlessand thus no shear stress can developon the
411)))
412 Soil Mechanics: Basic Concepts and Engineering Applications

Figure 10.1. Failure mechanism for frictionless strip footing.

contact area. This mechanism with some modifications has become the basic tool for evalua-
tion of the ultimate bearing capacity. The mechanism is comprised of three zones. In zone 1,
immediately the under
footing the major principal stress qu is in vertical direction. The minor
principal is horizontal
stress and can be found (in terms of qu) from Equation 4,14. In zone 3
the minor principal stress is equal to the overburden pressure yD and the major principal stress
can be found from Equation 4.13. In both zones the states of stress are assumed to be uniform.
The state of stress at zone 2 obeys a logarithmic function, which satisfies the equilibrium,
boundary conditions and the failure criterion, thus creating a relationship between two states
of stress of zone 1 and zone 3, [n this zone the soil is in plastic equilibrium and failure planes
passs trough point a'. Another set of faiIure surfaces are parallel to the Iogarithmic spiraI bc.
For a weightless soil (in the zones 1 to 3) it can be shown that the ultimate bearing capacity is:

= (10.1)
qu c'Nc +'tDNq
2 \302\260 = cot - (10,2)
O'(Nq 1)
= 2), N c
Nq exp(\177 tan \177') tan (45 + \177,/

where
Nq
and Nc are bearing capacity factors. For cu, 0u
= 0 soil Nq = 1, Ne =
(n + 2) = 5.14:
=
qu 5\"14cu +yD (i0.3)
The magnitude ofqu at the ground surface is 5.14Cu.
Example 10.1

Determine the ultimate bearing capacity of a frictionless strip footing of width 1.5 m at a depth

of 1 m in a soil with c' = 10 kPa, \177' -- 28 \302\260,

cu
= 105 kPa,
Ou=0 and g
= 19 kN/m 3.

Solution:
\"/D
= i9.0 x 1.0 = 19.0kPa for both drained and undrained conditions. Using Equations 10.2:
man 280\302\260 2 \302\260 \302\260 = 14.72, = cot - =
= e tan (45.0 + 28.0 / 2) N
c 28.0\302\260(14.72 1) 25.80.
Nq

Thus from Equation 10. I: qu

= 10.0x 25.80 + 19.0x 14.72 = 538kPa,
For the undrained conditions use Equation 10.3: = (n + 2)105.0+ 19.0= 559 kPa.)))
qu
 Bearing Capacity of Shallow Foundations and Piles 413

t
d

Figure 10.2. Failure mechanism for Terzaghrs bearing capacity solution.

10.2.2 Terzaghi's bearingcapacity equations

Terzaghi (1943) improved the solution

Prandtl to include the roughness of the footing, and the
weight of the failure zone. The failure mechanism for a general shear failure in a c', \177b'soil is

shown in Figure 10.2. The ultimate bearing capacity for a strip, square and circular footing and
the corresponding bearing capacity factors are:

qu
= c'N,: + + (10.4)
yDNq 0.5ByN\302\245

qu
= f.3c'N c + yDNq +
0.4ByN$ (lO.5)

q, = 1.2c'Nc + \177[DNq +
0.3ByN? (10.6)

tan \177\"
e(3n/2-\177') ' kPl\302\242
, N , = 0.5 -i\177\177 (10.7)
Nq- \302\260 c =cot\177'(Nq-l) Ny
tan\177
2cos2(45 +\177)'/2) (cos2 \177\"

The parameter is evaluated

kp,\302\242
from the equilibrium of the failure mechanism. From the given

10.3. Failure mechanism for solution in c,, = 0 soil.)))

Figure Terzaghrs bearing capacity \177,,
414 Soit Mechanics: Basic Concepts and Eng(neering Applications

1000

100

I0

0.1 I

0 10 20 30 40 50
\177'(degrees)

Figure 10.4. Bearing capacity factors using Terzaghi's equations.

values of N\177,
the following matching empirical equation may be proposed:

= '2
kp\177
(8q\177 -4\177'+ 3.8) tan2(60\302\260+ 0\"/2) (i0.8)

where qb' (in the first term) is ilt radians. In the undrained conditions with cu and 9u = 0, the
mechanism of Figure 10.3 yields Nc
= (3 / 2rr + 1) = 5.71,Nq = 1 and
N\177,
= 0. Figure 10.4
shows the variation of the bearing capacity factors with the effective internal friction angle.
Example 10.2
Re-do Example [0.t using Terzaghi's bearing capacity factors.
Solution:

Equations 10.8 and 10.7: =45.68, = 17.81, Nc

= 31.61 and = 15.31.
Usi\177tg kF\302\242 Nq N\177.

qu =c'Nc
= 10.Ox
31.61
+ 19.0x17.81
-- 0.5xl.5xlg.0x15.31 = 873 kPa.
+yDNq +0.5ByNy
For the undrained conditions No= 5.71, Nq= 1, and N\177,= 0, thus:

qu = 5.71cu yD + = 5.71x 105.0

+ 19.0x 1.0 = 619 kPa.

Example 10.3
Determine the ultimate bearing capacity of a squarefooting 1.5 m, at a depth of 1 m in a soil
with: c' = 10 kPa, 0\"
= 28\302\260,
q,=105 kPa, Ou=O, and5' = 19 kN/m 3.

Solution:

yD = 19.0 x 1.0=19.0kPa for both drained and undrained conditions. From Example 10.2:)))
 Bearing Capacityof Shallow Foundations and Piles 415

Nq
= 17.81, Ne = 31.61 and N\177,= 15.31. Using Equation 10.5:
qu
= 1.3x 10.0x 31.61 + 19.0x17.81 + 0.4xl .5 x 19.0x 15.31= 924 kPa.

For undrained conditions, No= 5.71, Nq= 1, and N\177,= 0, thus:

qu
= 1.3\303\227105.0x 5.71 + 19.0\303\2271.0 = 798 kPa.

10.2.3 Meyerhoj\302\260s bearing capacity equations

Meyerhof (1951, 1953,1963,1965and 1976) developed the bearing capacity equations by ex-
tending the Terzaghi's mechanism to the above the base of the footing
soil for both shaIlow
and deep foundations, where in the latter a local shear failure was also considered. The bearing
capacity factors N,, and Nq are identical analysis (Equations 10.2). Furthermore
to Parandtl the

shape of the footing, inclination of the applied load and the depth of the footing were taken
into account by introducing the corresponding factors of s, i, and d. For a rectangular footing
of L by B (L > B) and inclined load:

=
qu c'Ncscicdc + yDNqsqiqdq +
O.5ByN,\177s,tiTd \177 (10.9)

For vertical load: ic TM iq= i.i,= 1 and:

qu
= c'NcScdc + + (10.10)
yDNqSqdq 0.5ByN\177ts\177d 3,

The bearing capacity factor N\177,

is defined by:
= -
N,( (Nq 1)tan(1.40\") (10.11)

The bearing capacity factors are graphically presented in Figure 10.5. The shape, inclination
and depth factors are according to:

sc =l+0.2(B/L)tan2(45 \302\260+0'/2),sq=s,t=l+O.l(B/L)tan2(45\302\260+\177'/2)
(10.12)

For c,, qbu

= 0 soil s = s, = 1.
q t
ic = = 2,
(1-\177\302\260/90\302\260)
= (1-
\177/\177')
2
(10.13)
iq i5,

For cu, = 0 soil = 1.

qbu
i\177,

dc
=dv =l+O.l(DlB)tan(45\177+O'/2)
=i+0.2(D/B)tan(45\302\260+q/12),dq (10.14)
For cu, #u = 0 soil = = 1. The equivalent plane strain #' is related to triaxiai 9' by:
dq d\177,

= L) (10.15)
*'ps O\177ri(1.I-O.1B/
For the eccentric load the length and width of the footing rectangle are modified to:
L'= L-2eL, B'= B-2eB (10.16)

where eL and eB representthe eccentricity along the appropriate directions.

Example 10.4
Re-do example 10.3 using Meyerhof's bearing capacity factors.)))
416 Soil Mechanics: Basic Concepts and Engineering Applications

1000

IO0

0.1
0 10 20 30 40 50
e' (degrees)
Figure 10.5. Bearing capacity factors using Meyerhof's equations.

Solution:
From 10.2 = 14.72, Nc
= 25.80
(Example i0.1), Equation 10.11:
Equations Nq using

=
(/4.72
- 1)tart(1.4\303\227 28.0 = 11.19.
\302\260) The load is vertical, thus = = = 1.
i\177. iq iy
N\177,

Calculate shape factors from Equations 10.12 and depth factors from Equations 10.14:
= = 1 + 0.1(1.5/1.5) 2 \302\260
tan (45.0 + 28.0\302\260/2) = 1.28,
Sq s\177,
sc = 1 + 0.2(i.5/1.5) tan2(45.0\302\260 + 28.0\302\260/2) = 1.55.

d r = l + 0.2(1.0/
\302\260
1.5)tan(45.0 + 28.0\302\260/2) = 1.22,

= = 1+ 0.1(1.0/1.5) tan(45.0
\302\260
+ 28.0\302\260/2) = 1.11. From Equation 10.10:
dq dy

q, = 10.0\303\22725.80\303\2271.55\303\2271.22+ 19.0\303\22714.72\303\2271.28\303\2271.i 1+ 0.5 \303\227

1.5\303\227 1.28\303\227
i9\303\22711.19\303\227 1.i1,

qu
= l112 kPa. For the undrained conditions, No= 5.I4, Nq=l, and 0.
N\177,=
= 1 + 0.2(1.5/1.5) tan 2 \302\260 = 1.2, = = 1.0,
s,. (45.0 + 0/2)
Sq sy
d c = 1+ 0.2(1.0 / 1.5) tan(45.0
\302\260
+ 0 / 2) = 1.13,
dq
=
d\177t
= 1.0,

= 105.0\303\2275.14\303\2271.2\303\2271.13 + 19.0x = 751

1.0\303\2271.0\303\2271.0 kPa.
q,

Example 10.5
Re-doexample 10.4 assuming that the centralload appliedto the footing is inclined at an an-
gle \177t= 10
\302\260
to the vertical. The load is locatedin a plane parallel to the side of the square foot-
ing. What is the ultimate value of this load?
Solution:)))
 Bearing Capacity of Shallow Foundations and Piles 417

1000

100

\"\"
\177
10 Nc- \177\" \"\177'\"

'
t

0.1

0 10 20 30 40 5O
\177'(degrees)

Figure 10.6. Bearing capacity factors using Hansen's equations.

From Equations [0.13: i,.

= (1 = - 10.0\302\260/90.0\302\260)
2 =
0.79, = (1 - 2 =
t0.0\302\260/28,0\302\260) 0.41.
iq iy

The shape and depth factors are the same as in Example 10.4. From Equation 10.9:
qu
= 10.0x 25.80x 1.55 x 1.22x0.79+ 19.0x 14.72x 1.28 x 1.11x0.79 +
0.5x1.519 x I 1.19 x 1.28x 1.1Ix0.41 792
x = kPa. The ultimate value of the vertical load is:
Qm,
= 792x1.52 = 1782 kN, Qu
= 1782/c\302\260s10.0\302\260 = 1810 kN.

For undrained conditions the shape and depth factors are identical to the values calculated in
Example 10.4. ic = iq =0.79 (as above) and iv= 1.0,thus:
qu
= 105.0x5.14x 1.2xl.13x0.79 + 19.0x1.0xl.0xl.0x0.79= 593 kPa.
Q,w
= 593x1-52 = 1334kN, Qu = 1334/c\302\260s10.0\302\260 = 1355 kN.

10.2.4 Hansen's bearingcapacity equations

Hansen (196i and 1970) extendedMeyerhof's solutions by considering the effects of sloping
ground surface and tilted base (Figure 10.7) as well as modification of N-\177and other factors.
For a rectangular footing of L by B (L > B) and inclined ground surface, base and load:

qu
=
c'Ncscicdcbcgc + yDNqsqiqdqbqgq +
0.5ByN,{s,{iyd\177bygy (10.17)

For horizontal ground surface and base bc = bq =

by
= gc =
gq = gy= I and the general bearing
capacity equation becomes the same as Equation 10.9. For the undrained conditions:

qu =5\"14Cu(l+sc +de -ic -be -gc)+Y D (10-18)

L)))
4 i8 Soil Mechanics: Basic Concepts and Engineering Applications

\177
T =.Vtan\177+ CbA

Figure 10.7. Identification of the terms in Hansen's bearing capacity equations.

The bearing capacityfactors of Nq and Nc are defined by Equations 10.2; N,\177

is defined by:
= - 1) tan (10.19)
Ny 1.5(Nq 9'

sc, B=l+NqBicB,
N L ' Sq'B '
=l+\177iqBsin\302\242\"
,
Sy'B =I-0.4\177BLi. \302\242
--
\"B>0.6 (10.20)
c

\"
=
Sc,L =l+NqLi e,L, Sq,L =l+__\177iq,Lsinc\177,L Sy, L ,L
>0.6 (10.21)
N
c B 1-0.4-\177i\177

Forcu, Ou=Osoil so, \177=0.2(B/L)ic, \177, so, L =0,2(LIB)it, L (10.22)

1-iq'i \" \" 0\"7I-li

= = (1 0\"5Hi \177xl
, = (1 (10.23)
ic'i iq'i -1' zq'i t-\177,i
V + Ac
Nq

where i (in Equations 10.23) =

B or L, 2 _< at -< 5, 2 _<
a2
< 5, A is the area of the footing base
and cb is the cohesion mobilized in the footing-soil contact area. For the tilted base:

(0.7-rl\302\260/450\302\260)Hi
i\177,,i
(10.24)
V H-\177cb c--\177
=[1 .j
= 0.5-
For cu,
= 0
\302\242\177u soil
ic,i 41-'Hi / Ac b i =B or L (10.25)
In the above equations B and L may be replaced by their effective values expressed by Equa-
tions 10.16. The depth factors are specified in two sets.
For D/B<I, D/L<_I:

de,\177=I+0.4D/ B, dq, B =l+ 2tan\177'(1-sin#')2D/ B (10.26)

de,L =I+0.4D/ L, dq, L =l+2tan\177'(1-sin\177')2 D/ L (10.27))))

 Bearing Capacityof Shallow Foundations and Piles 419

For D/B>I, D/L>I:

= 1 +0.4tan -1 2
de, B (D/B), dq,B = 1+ 2
tanO'(1 -sin\177') tan-l(DlB) (10.28)
= I + = 1`+2tang\"(I 2
de.L 0.4tan I(DIL), tan-l(DlL) (1.0.29)
dq, L -sin\177')

For both sets

d\177,
= 1..0
(10.30)

For cu, \177u

= 0 soil de. B = 0.4(D/B), de,L = 0.4(D/L) (10.31)
For the sloping ground and tilted base the ground factor gi and base factor bi are proposed by
Equations 10.32 to 10.35. The angles [3 and rl are at the same plane either parallel to B or L:

5
ge = 1-[3\302\260/147\302\260,
gq
=
g\177,
=
(1-0.5 tan[3) (10.32)

For c,, 0\177J= 0 soil gc =[\177\302\260/147\302\260 (10.33)

bc = 1- / 147 \302\260,
\"q\302\260
= e-2q tan
0\", = e-27\1771
tan ,'
(10.34)
bq by

For c,, \177u

= 0 soil bc = rl\302\260/147
\302\260
(10.35)

Example 10.6

Re-do example 10.4using Hansen's bearing capacity factors.

Solution:
\177
From Equations 10.1.9:
N\177.
= 1.5(1.4.72 -1.0) tan 28.0 = 10.94.
Shape factorsare calculatedfrom Equations 10.20 by setting the inclination factors to 1.0.
- 1+ (1.5/1.5)sin28.0
s,. = 1+(1.4.72/25.80)(1.5
\302\260
/ 1.5) = 1.57, = 1.47,
Sq
= = 0.6.From Equations 10.26 and 10.30:
s,/ 1-0.4(1.5/1.5)
= =
d c 1+0.4(I.0/I.5) 1.27, = I + 2 tan 28.0\302\260(1- sin28.0\302\260)2 (1.0/1.5) = 1.20 and = 1.0.
dq d\177.

Using Equation 1.0.17:

= 10.0x 25.80\303\2271`.57 1.27
\303\227 + 19.0\303\22714.72xl.47x1.20+ 1.5 \303\22719x1`0.94\303\2270.6x
0.5\303\227 1.0,
qu

q\177
= 1101 kPa. For the undrained conditions Nv = 0. The shapefactors sq
and s,/are calculated
from the same equations for 0' > 0 case,but sc is calculated from Equations 1.0.22, thus:
= 0.2,Sq=
s c - 0.2(1..5/1.5) 1.0, s\177.=
0.6.

Using Equation 10.31 for dc (0' = 0) and dq dy: Equations 10.26 for and

dc = -
0.4(1.011.5) 0.27,dq = 1.0, d\177,= 1.0. Substituting in Equation 10.18:

q, =5.14x105.0(1 +0.2 +0.27) + 19.0=812kPa. The resul\177s of Examples 1.0.3, 10.4 and 10.6
are tabulated below:

Soil condition Terzaghi Meyerhof Hansen

\302\260
q\177,(kPa):c'= lO.OkPa, q\177'=28 921 1112 ll01
qu (kPa): c,, = 105 kPa. \177u= 0 797 751 812)))
420 Soil Mechanics: Basic Concepts and Engineering Applications

10.2.5 VesiUs bearing capacity equations

Vesi6(1973)suggested for ultimate bearing capacity that is similar to Hansen's
a procedure

method with minor and some selected factors. This method seems easierto
modifications in Ny
use, as there is no interrelationships between different factors. In this method Equations 10.17
and 10.18 are used with the following modifications:
= 2(N + 1) tan
Ny q 0\" (t0.36)

= 0
For cu, 0\177 soil with the sloping ground (Figure 10.7):
N = -2sin 13 (10.37)
v

The other terms are as foIlows:

sc, o=l+(Nq/Nc)(B/L),sq, o=l+(B/L)tan(\177',Sy, B=l-O.4(B/L)>_0.6 (10.38)

so,L =I+(Nq/Nc)(LIB),sq, L =l+(L/B)tand/,s\177, L =1-0.4(L/B)>_0.6 (10.39)

For cu, \177u

= 0 soil Sc,B = 0.2(B/L), So, L
= 0.2(L/B) (10.40)
For ic, i use Hansen's equation; other terms are definedby:

= (1 mi+l i = B\302\260rL o(10.41)

iq,i=(l V+ A cH,\177cot\177,)
mi, iv, i
v+AHc;cotd\177,)

For cu, \177u

= 0 soil ic,i =
1-mill i/A%N
=
c, i B or L (10.42)
m
B
= (2+B/L)/(1+ B/L), mL =
(2+ L/ B)/(I + L/ B) (10.43)
For the sloping ground and tilted base (Figure 10.7):
= = - 2
gc iq
-(1 -iq)/(5.14tan\177'), gq g,/= (1 tartlY) (10,44)

For cu, 9u
= 0 soil ge = [\177/5.14 (10.45)

= 1- = = (!-rl 2
bc 213 /(5,14 tan \177') , tan\177') (10.46)
bq b\177,

For c,, \177?u

= 0 soil bc=13/5.14 (10.47)
The depth factors are the same as Hansen'smethod. For circular bases in both Hansen's and
Vesid's methods the dimensions of an equivalent square may be used,
Example i0.7
A square footing of 1.5 m is to be constructedin sand with c' = 0, \177' = 40 \302\260.
The thickness of
the footing is 0.45 m and its top surface is level with the horizontal ground surface. The foot-
ing is subjected to a central vertical force of 700 kN and a central horizontal force (parallel to
the side) of 210 kN. Find the ultimate bearing capacity by the (a) Meyerhof's method, (b)
Hansen'smethod and (c) Vesi6's method. The unit weight of the sand is 18 kN/m 3.)))
 Bearing Capacity of Shallow Foundations and Piles 421

Solution:

= 0.45\303\22718.0 = 8.1 kPa. Find Nq from Equation 10.2:

,/D
\302\260
2 n tan 40.0 2
(45 + 9' / 2) = e
= \302\260 \302\260 \302\260 = 64.19.
exp(\177 tan \177)')tan tan (45.0 + 40.0 / 2)
Nq
(a) Meyerhof'smethod: find N\177from Equation 10.11:
=( - = -
1) tan(1.40' ) (64.19 1)tan(1.4 x 40.0 = 93.68.
\302\260)
N,t Nq

Using Equations 10.12 and 10.14:

L) tan (45 + 9' / 2) = 1+ 0.1(1.5/1.5)tan
= = 1 + 0. I(B / 2 \302\260 2 \302\260 \302\260 = 1.46.
(45.0 + 40.0 / 2)
Sq s\177,

= = 1+ 0. I(D / B) tan(45 \302\260

+ 9'/2) = 1.0 + 0.1(0.45 / 1.5)tan(45.0 \302\260
+ 40.0
\302\260 =
t 2) 1.06.
dq d,\177
-1 =
Inclination angle ct = tan (210.0 / 700.0) 16.7 thus from Equations
\302\260, 10.13:
2 :
= (1-ct\302\260/90\302\260) 2 = 2
(1-16.7\302\260/90.0\302\260) =0.66, (1-16.7\302\260/40.0\302\260) =0.34.
iq i\177,

Using Equation 10.9:

qu
= 8.1\303\22764.19xl.46xl.06\303\2270.66+0.5\303\227l.5\303\22718.0x93.68xl.46xl.06x0.34 =1196 kPa.
(b) Hansen'smethod: find N v from Equation 10.19:
= - - \302\260
N3, 1.5(Nq 1) q\177' 1-5(64-19
= =
1.0)tan tan 40.0 79.53.

The inclination factors are calculatedfrom Equations 10.23 with cq = c\1772= 3.

= [I-0.5Hi/(V + Ac \1771 = 3 =
iq b cot9\]") [1-0.5\303\227210.0/(700.0+0.0)] 0,61,
= [1 - - 3 =
0.7H i/(V + [1 0.7 x 210.0/(700.0 0.49.
Ac c\1772 = + 0.0)]
i3, b cot0')]

CaIculate shapeand depth factors from Equations 10.20, 10.26and 10.30:

=l +(B/
sq L)iqsing'=l +(l.S/l.5)xO.61\303\227sin40.0\302\260 =1.39,

- = 1 - 0.4(1.5 / 1.5)x 0.49= 0.80.

s,\177
= 1
0.4(B / L)i,\302\242
= 1+ 2 tan \177'(1 -sing')2(D/B)
= 1+ 2 tan 40.0\302\260(1
- sin = 1.06, =
40.0\302\260)2(0.45/1.5) 1.0.
dq d\177,

Thus from 10.17 and with c\" = 0 and bc = bq = = gc = = = 1:

Equation by gq g),
= = 8.1x64.19xl.39\303\2270,61xl.06+
qu yDNqsqiqdq +O.5ByN,ts@td, t
0.5x 1,5\303\227
18.0 x 79,53x 0,Sx 0.49 x 1.0= 888kPa.
(c) Vesid's method: from Equation 10.36:

= = + 1)tan 40.0 \302\260

2(Nq + 1) 0' 2(64.19
tan = 109.40.
Ny
Shape functions are calculatedfrom Equations 10,38:
9' = 1+ (1.5/1.5) tan 40,0 = 1.84, = 1-0.4(1,5/1.5) = 0.60.
= 1+ (B/L)tan \302\260 = 1-
Sq s\177.
0.4(B/L)
Depth factors are the same as Hansen'smethod: dq- 1.06, dr= 1.0, From Equation 10.43:

tn = (2 + B/ L)/(1+ B/ L) = (2 + / 1.5)/(1 + 1.5 /1.5) = 1.5, using Equations

1.5 10,41:
= [1- H I(V + Ac cot\177')] m = [1 - 210.0/(700.0 + 0.0)]15 = 0.58,
iq i b
- m+l =
i\177\302\242
[1- H i/(V + Ac b cot\177')] [1- 210,0/(700.0 + 0.0)]
2.5
= 0.41, thus:

q, = yDNqsqiqdq +0.5ByNvsviyd \177

= 8.1x64,19xl.84x0.58xl.06+

0.5 \303\227
1,5\303\22718.0x109.40\303\2270,6\303\2270.41 xl.0 =951 kPa.)))
422 Soil Mechanics: Basic Concepts and Engineering Applications

10.2.6 Skempton's Nc value

For c,, \177u= 0 soil Skempton (1951) suggestedthe following formulas:

= + yD
qu cu Nc (i0,48)

= 5.14(1+ 0.2B/L)(I+ - 0.03ID2 e

Nc 0.24D / B / B ) (10.49)
For circuIar baseB = L = diameter of the footing. The values of No for strip and square (or cir-
cular) footing are presented graphicaIly in Figure 10.8.

Example i0.8
Determine the ultimate bearing of a square footing 1.2 m, at a depth of 2.5 m in a soiI
capacity
with cu = 105 kPa, Ou=0, \302\245
= 19 kN/m \".

Solution:

yD = 19.0 x 2.5 = 47.5\"kPa. Using Equation 10.49:

Nc = 5.14(1+ 0.2x1.2/1,2)(1 +0.24x2.5/1.2-0.031x2.52/1.22)
=
8.42, thus:

qu
= cuNc +yD = 105.0x8.42+
47.5 = 932 kPa.

10.2.7 Effect of water table on bearing capacity

At the presence of hydrostatic water an effective

weight ?e must be substituted
unit in tlae bear-

ing capacity expressions. The effective weight is

unit 50% to 100% of the unit weight of the
soil above the water table. The following equation is a reasonable estimation for \177'eunder the
footing Ievel:

1o I I i

Square or circle (B / L = 1)
9

Nc 7
J Strip (B / L
= O)

4
0 1 2 3 4
DIB
Figure 10.8. Nc values proposed by Skempton (]951).)))
 Bearing Capacity of Shallow Foundations and Piles 423

ye=V'+ z,,./B(Y-V ') (z\177 <_B) (10.50)

where zw is the depth of water level from the base, y is the unit weight above the water level
and g' is the submerged unit weight of the soil. If water table is on the foundation base (Zw
=

0), Ye = \177; for zw = B, Ye = Y. An alternative equation can be obtained by calculating an average

unit weight within the zone 1 of failure mechanism of Figure 10.1:

\177H-Zw.2] ,\177H -zw. 2

where H = Btan(45\302\260+\302\242'/2). Note that in most cases tile third term of the bearing capacity
expression (0.5 B y Ny) has not a significant value and its further reduction by water will have
little eff\177t on the bearing capacity. However in granular materials this te\177 is the mare con-
tributor (apa\177 from yDNq) and its reduction due to the water table could be of significam
value. For the buried %otings similar reduction Nctors in the fo\177 of effective unit weight are
applied for the surcharge load

Example 10.9
Re-do Example 10.7assuming that the water table is at the footing level (0.45 m below the

ground surface). The saturated unit weight is 21 \177/m \177.

Solution:

Using \177uation 10.50 or 10.51:

Ye the footing) = \177 = 21.0 -
(under 9.81= 1i. 19\177/m 3.

(a) Meyerbof's method

= = 0.66, and
Nq= 64.19.N\177 = 93.68, sq sv
=
1.46, dq = d\177
= 1.06,
iq
= 0.34.
i\177

From Equation 10.10:

qu
= 8.Ix 64.19x 1.46xl.06x0.66 1.06x0.34
+0.5x 1.5xI 1.19x93.68xl.46x = 945 \177a.

(b) Hansen's method

Nq
= 64\17719,
Ny
= 79.53,
sq
= 1.39,s\177= 0.80, dq
= 1.06, d
y
= 1.0,
iq
= 0.61, and i
7
= 0.49.

From \177uation 10.17:

q,
= 8.1 x 64.19x 1.39x1.06x0.61
+ 0.5x1.5xl 1.19x79.53x 0.80x 1.0x0.49 = 729\177a.
(c) Vesid's method
= 1.84, = 0.60, = 1.06, d = 1.0, = 0.41.
Nq= 64.19, NT= 109.40. Sq sy dq 7 iq =0.58, and i7

From Equation 10.17:

q,
= 8.1x64.19xl.84xl.06x0.58+0.5xl.5xl 1.19x109.40x0.60xl.0x0.41 = 814kPa.

10.2.8 Allowable bearing pressure and factor of safety

Allowable bearing (or soil) pressure qa is the contact pressure beneath the footing which (a)
satisfies the settlement requirements, (b) provides safety against shear failure. Thus the ulti-
anate bearing capacity must be divided by a desired factor of safety F to provide a safe bearing)))
424 Soil Mechanics: Basic Concepts and Engineering Applications

capacity equal to or greater than the allowable bearing pressure. The settlement requirements
depend on the type of the structure; however a maximum differential settlement (between ad-
jacent columns of the structure) of 25 mrn is accepted universally by geotechnical and struc-
tural engineers. For sands a maximum settlement of 40 mm for isolated footings and up to 65
mm for raft foundations is recommended. The corresponding maximum settlements for clay
are 65 mm and 100 mm respectively (MacDonald & Skempton, 1955; Walsh. 1981).
The magnitude of factor of safety applied to the ultimate bearing capacity depends on the
type of the foundation. For spreadfootings it is between 2 to 3, whilst for mat foundations it

may selected within the range 1.7 to 2.5 (Bowles, 1996). The factor of safety in terms of actual
contact pressure q <_ qa and footing depth D may be expressed by:
-
- qu yD
F (10.52)
q-vD
Most of the building codes recommend the Ultimate Strength Design (USD) for design and
analysis of foundations. In this method the working loads are factorised and combined (ac-
cording to the code) to give the ultimate load applied to the footing. This load in many cases is
about 1.5 times of the working higher or lower values may be obtained depend-
load; however
ing on and critical combination of the involved
the nature loads. Under this ultimate load the
footing material is in plastic state. While the initial plan section of a footing is estimated using
working loads and allowable bearing pressure, the cross section and reinforcements are calcu-
lated by USD method. The contact pressure under the ultimate load is called ultimate soil

pressure. If the allowable bearing pressure is obtained by application of a factor of safety of 3

to the ultimate bearing capacity, and the factorised ultimate load is 1.5 times of the working
load, then the ultimate soil pressure is 50% of the ultimate bearing capacity. This ensures the
factor of safety of 2 against shear failure of the foundation soil at the time when the foundation
material has reached to its ultimate strength.

Example i0. i0
Find the factor of safety against shear failure in Examples 10.7 and 10.9.

Solution:

Using Equation i0.52 for Example 10.7:

Note 7D = 8.1 kPa and,
q
= (contact
pressure) = 700 (vertical load) ! 1.52 = 3 t 1.1 kPa.
(a) Meyerhof's method: F = (1196 - 8. 8.1)
= 3.92, -
i)/(311.1
(b) = -
Hansen's method: F (888 8.1)1(311.1 - 8.1) 2.90,
=

(c) Vesid's method: F =(951-8.1)1(311.1-8.1) = 3.11.

For the Example 10.9:
(a) Meyerhof's method: F = (945 - 8.1)1(31i.1- 8.t) =
3.09,

(b) Hansen's method: F - (729-8.1)/(31

i.I-8.1)= 2.38,
(c) Vesid's method: F = (814-8.1)I(311.1-8.1)
=2.66.)))
424 Soil Mechanics: Basic Concepts and Engineering Applications

capacity equal to or greater than

allowable bearing pressure. The settlement
the requirements
depend on the type of however a maximum differential
the structure; settlement (between ad-
jacent columns of the structure) of 25 mm is accepted universally by geotechnical and struc-
tural engineers. For sands a maximum settlement of 40 mm for isolated footings and up to 65
mm for raft foundations is recommended. The corresponding maximum settlements for clay
are 65 mm and 100 mm respectively (MacDonald & Skempton, 1955; Walsh, 1981).
The magnitude of factor of safety applied to the ultimate bearing capacity depends on the
type of the foundation. For spreadfootings it is between 2 to 3, whilst for mat foundations it
may selected within the range 1.7 to 2.5 (Bowles, 1996). The factor of safety in terms of actual
contact pressureq <- qa and footing depth D may be expressed by:

-yD
F =
q\"
(10.52)
q-yD
Most of the building codes recommend the Ultimate Design
Strength (USD) for design and
analysis of foundations. In this method the working are factorised and combined
loads (ac-
cording to the code) to give the ultimate load applied to the footing. This load in many cases is
about 1.5 times of the working load; however higher or lower values may be obtained depend-
ing on the nature and critical combination of the involved loads. Under this ultimate load the
footing material is in plastic state. While the initial plan section of a footing is estimated using

working loads and allowable bearing pressure, the cross section and reinforcements are calcu-
lated by USD method. The contact pressure under the ultimate load is called ultimate soil

pressure. If the allowable bearing pressure is obtained by application of a factor of safety of 3

to the ultimate bearing capacity, and the factorised ultimate load is 1.5 times of the working
load, then the ultimate soil pressure is 50% of the ultimate bearing capacity. This ensures the
factor of safety of 2 against shear failure of the foundation soil at the time when the foundation
material has reached to its ultimate strength.

Example 10.10
Find the factor of safety against shear failure in Examples 10.7 and 10.9.

Solution:

Using Equation 10.52for Example 10.7:

Note yD
= 8.1 kPa and,

q
= (contact
pressure) = 700 (vertical load) / 1.52 = 31 i. 1 kPa.
(a) Meyerhofs method: F = (i 196- - 8.
8.1)/(311.1 i)
=
3.92,

(b) Hansen's method: F =(888 -8.1)/(311.1-8.1)- 2.90,

(c) Vesid's method: F = (951-8.1)/(311.1-8.1) = 3.11.
For the Example 10.9:
(a) Meyerhofs method: F = - 8.1)/(311.1- 8.1)= 3.09,
(945
(b) Hansen's method: F = (729-8.1)/(311.I-8.1) = 2.38,
(c) Vesid's method: F -
(814- 8.1)/(311.1- 8.1)= 2.66.)))
 Bearing Capacity of Shallow Foundations and Piles 425

10.3 FIELD TESTS

10.3.i Plate loading test

The plate-loading test is carried out

to estimate the bearing capacity and settlement beneath the
single footings (Section A loading plate is circular or square
5.4.6). in shape and is manufac-
tured from machined steel plate with a minimum thickness of 25 rnm. The diameter of circular
2
plates ranges from 150 mm to 760 ram. The square or circular plate with 1 ft area has many

practical applications in geotechnical site investigations. The load is either of gravity type ap-
plied through a platform or reaction type applied by a hydraulic jack. The test is carried out in
a pit with its bottom located at the actual footing level. The diameter of the pit is 4 to 5 times
of the diameter of the plate to allow the developments of failure planes to the ground surface
(bottom of the pit). The toad is applied in increments each 1 / 5 to 1 / i0 of the predicted ulti-
mate load. For each increment a time (logarithmic scale)-settlement plot is constructed from
the recorded data. The next increment is applied after a specified time interval (not less than 1
hour) when no settlement is observed. The settlement-load plot is also constructed to establish
the maximum contact pressure or the contact pressure needed for a specified settlement (com-

monly 25 ram). Equation 5.63 can be used to relate the test results to the actual tooting; how-
ever the disadvantages of an extrapolation has been investigated by some authors (Section
5.4.6) and must not be used if the ratio of the width of the footing (B) to the width or diameter
of the plate exceeds 3. In sands a screw-plate test is performed without a need for a pit excava-

tion, The plate loading test procedure may be found in ASTM D-1194, D-1195,and D-1196.

10.3.2 Standard penetration test

Standard Penetration Test (SPT) is a dynamic in-situ test

performed mostly in cohesionless
soils to estimate the allowable bearing pressure as well internal friction angle and the ul-
as the
timate bearing capacity. Furthermore the correlation of SPT results and elastic properties of
soils has given useful practical equations (Section 5.4.7). In the absence of a sampling tech-
nique to provide undisturbed specimens of cohesionlcsssoilsSPT is considered a simple but
efficient tool in site
investigation of these soils. This test determines the resistance of the soil
to the penetration of a split barrel sampler and obtains disturbed soil specimens for identifica-
tion purposes. The details of basic equipment and test procedure may be found in ASTM D-
1586,BS 1377and AS 1289.6.3.1. The schematic view of the sampler is shown in Figure
10.9. The section is comprised of a driving shoe, a split barrel for collection of the disturbed
sample and a sampler head for connection to the rod and drilling assembly. The central part of
the sampler is of split type construction (to allow the removal of the recovered soil) with pro-
vision for a liner. The sampler is driven 450 mm through a bottom of a borehole(already
drilled in the ground) using a hammer of 64 kg falling freely from a vertical distance of 760
ram. The number of blows are counted for each 150mm penetration. The sum of number of

blows for the last two 150 mm penetrations gives the SPT number N (also called penetration
resistance). The SPT number for adjacem holes from equipment made by different manufac-
turers are not equal and must be normalized to an energy ratio recommended by the code or)))
426 Soil Mechanics: Basic Concepts and Engineering Applications

Ddving shoe Sampler head

I _

Split barrel
\177-I Four vents 13 mm Dia.

, ......

20 mm 20 mm
465 mm 125 mm

650 mm
\177-
t

Figure 10.9. Schematic diagram of a split barrel sampler.

computed by the user (Seed etal., 1984; Skempton, 1986). The energy ratio is defined by:
Er =(Eac/Ein)xlO0% (10.53)

where Ein is the input driving energy caused by 760 mm free dropping of 64 kg hammer

(0.475 kN.m) and Eac is the actual energy transferred to the sampler. The energy ratio may
change from 50% to 80% and more depending on the type of the hammer. In some countries
this ratio is standardized. For example in UK for rope-pulley or cathead lift hammer Er is
50%, whilst for hammers with automatic trip it is 60%. A corrected N value for an apparatus
of E,-I with N1 count can be calculatedfrom the following equation:
Er xN =
Er! \303\227N
1 (10.54)

where Er is the standardized energy ratio. Note that to establish Erl we need to calibrate the
apparatus (drilling rigs) on regular bases. Further adjustments are due to water table, overbur-
den pressure, length rod (reduction effects for lengths
of the drilling less than 10 m), sampler
liner, and The correction for water table is carried out using the following
borehoIe diameter.
equation (Peck, etal., 1974)where he SPTnumber N must be multiplied by parameter Cw:

c,,, = 0.511
+ D\177./(D + B)] 00.55)

where Dw and the depths

D are of water table and footing base from the surface respectively.
In the current the correction for water table is ruled out due to conservative
practice \177ature of
the above equation and that the SPT number obtained below the water table reflects the reduc-
tion in the N value. The early experimental results by Gibbs & Holtz (1957) and Meyerhof
(I 956) showed that the increase in overburden pressure influences the SPT number; thus the
measured field N value must be multiplied by a correction factor CN. In the Meyerhofs pro-
posal the corrected N value depends on the square of the density index 1o (or relative density:
Equation I.. 13) and changes li nearly with effective overburden pressure P'o accord ing:)))
 Bearing Capacity of Shallow Foundations and Piles 427

N = \177v(a+bP'o) (10.56)

Terzaghi et al. (1996) proposedthe following table for sands:

Table 10.1.Relationship between SPT number and density index (Terzaghi et al., 1996),
Sand SPT number N

Very loose 0-4

Loose 4- i 0
Medium 10-30
Dense 30-50
Very dense Over 50

Peck et al. (1974)presented a correction factor that has a considerable applications in practise:

CN = (P'o in
0.771og(2000/p\177,) kPa) (10.57)

Based on Meyerhof' s concept,Skempton (1986) proposed the following equation:

CN =
a/(b+ P'o) (10.58)

wherep'o is in (kPa); a and b are according to the table below:

Table 10.2.Parameters to obtain the correction factor CN (Skempton, 1986).

Sands Rangeoflo a b

Fine-medium 0.35 to0.65 200 100

Dense-coarse 0.65 to 0.85 300 200
Overconsolidated 0.85 to i.00 ]70 70

Correctionfor overburden pressure may be carried out by the following equation, which cov-
ers most of the proposed corrections by different authors (Liao & Whitman, 1986):

= (10.59)
CN 495'76/Po

wherep'ois the effective overburden pressure at the footing level in kPa,

Correction factors for the drill rod length and borehole diameter are given by Skempton
(1986). For rods longer than 10 m no correction is necessary whilst for rods less than 10 m a

reduction factor of 0,75 (short rods) to 1 (10 m rod) is applied. For boreholes65 mm to 115
mm no correction factor is needed.For wider boreholes the correction factor is greater than 1
being 1.i5 for 200 mm borehole, The existence of a liner in the split barrel increases the N

value due to side friction, tiros a reduction factor must be applied. The amounts of this factor
for dense and loose materials are suggested 0.8 and 0.9 respectively.

The correlationsbetween shear strength parameters and SPT number N have been reported

by number of authors. Peck et al, (1974) gave correlationsbetween corrected N (for effective

overburden pressure) and the bearing capacity factors, Furthermore the correlations were ex-
tended to estimate the 4' values of the normally consolidated sands (Schmertmann, 1975) and)))
428 Soil Mechanics.\" Basic Concepts and Engineering Applications

overconsolidated clays (Stroud, i989). Table 10.3relates the SPT number to the consistency
states of clayey soils (Terzaghi et al., 1996).While using these tables one must take into ac-
count the energy ratio corresponding to SPT number.

Table 10.3. Bearing capacity in clay soils and the corresponding SPT numbers (Terzaghi et al., 1996).

Consistency Very soft Soft Medium Stiff Very stiff Hard

SPTnumberN <2 2-4 4-8 8-15 15-30 >30
q,, (kPa) < 25 25-50 50-100 100-200 200-400 > 400

Correlationbetween SPT number and allowable bearing pressure was investigated by Terzaghi
& Peck (1967) for footings on sands
relating the isobars of SPT number N to allowable bear-
ing pressure qa and width of the footing B. Howeverit has been shown that these charts are

very conservative being 50% or more on the safe side. Meyerhof (1956 and 1974) provided
the following relationships:

- 12Se\177
kd B< 1.22 m
qa (10.60)
25
8SeN
--
qa
= (B +O.305)Z
\"
kd
B > 1.22 m (10.61)
25 B
where Seis the elastic settlement of the layer in ram, \177 is the average SPT number within the
influence depth of D (depth of the base) to D + B and ka is defined by:

ka = i + 0.33D/B (10.62)
Bowles(1996)improved Meyerhof's equations by 50% increase in the allowable bearing pres-
sure according:

- 20SENk a B < 1.22 m (10.63)

qa 25

- 12.5SEN
B> 1.22 m
qa ( B+O'305)2k a (10.64)
25 \177
Figure 10.10 the allowable bearing
shows pressure corresponding to 25 mm settlement con-

structed above equations for D/ B = 0. In the above equations an energy ratio

from the of 55%
must be assumed. Burland, & Burbidge (i985) suggested the following equations for sands
(Section 5.4.6,Equations 5.69 to 5.71). For normally consolidated sands:

Se\177l'4
qa = (10.65)
1.7B0.75
For overconsolidated sands:

qa \177-
(qa > Pc) and =
(qa < P\177) (10,66))))
1.7B 075
Pe qa
\177 1.7B 075
 Bearing Capacity of Shallow Foundations and Piles 429

ooo
\177 I I
goo
,, DIB\177O

\177
\177 \177. -._....\177 N=50
700 , \177.\177
1.2.2m ',
600

5OO
t \177\177 = 30
N
400

200
\177\177 N=ln
100
\177
0 IN=5

0 1 2 3 4 5 6
B (m)
Figure 10. I0. Allowable bearing pressure corresponding to 25 mm settlement (Bowles, 1996).

where p'c is the The magnitude of \177 may be corrected according to

preconsolidation pressure.
Equations 5.67 and 5.68 (tbr and gravelly sands respectively) including
saturated sands the
various adjustments mentioned above. For rectangular footings the assumed settlement must
satisfy the shape factor expressed by Equation 5.72. This means the allowable bearing pressure
(including the term 2/3p,\177) must be reduced by the shape factor (maximum 1.56). If an in-
compressible layer is located within the influence depth Zz, the calculated qa (including the
term 2/3pf.) may be increased by the correction factor defined by Equation 5.73.

Example i0.11
For the SPTresults given 5.21 and
in Example a square footing with B = 2 m calculate the al-
lowable bearing pressure corresponding to 25 nma settlement (a) if the base of the footing is on
the ground surface, (b) if the base of the footing is 1 m below the ground surface.

Solution:

Meyerhof's method: from Example 5.21the average uncorrected (for water table) SPT values
for cases (a) and (b) are 15(z = 0 to z = 2.0 m) and 20 (z = 1.0 m to z = 3.0 m) respectively.

Case (a): qa - 8SeN(.B+O305)2k- 8\303\227250\303\22715(20+0305):\303\2271.0 = 159.4kPa,

d
25 B 25 2.0

- 8x25.0x20
Case (b): qa ( 2.0+0.305)2 x(i+0.33 1.0) = 247.6kPa.
25 2.0 2.0
Using Bowles method the qa values obtained from Meyerhof's method are multiplied by the
ratio of 12.5 / 8. For case (a) qa = 249.0 kPa and for case (b) qa = 386.9 kPa.)))
430 Soil Mechanics: Basic Concepts and Engineering Applications

Burland, & Burbidge method: the average corrected value of SPT within the influence depth is
14 and cases (a) and (b) respectively.
17 for

\177'\177 k4 \1774

case
_ Se\177 _ 25-0x14
-351.8 case
_ 25.0x17 = 461.'7kPa.
(a) qa kPa, (b) qa
t.7B0.75 1.7X2.00.75 1.7x 2.00.75

10.3.3 Cone penetration test

Cone Penetration Test (CPT) is a static in-situ for subsurface exploration in fine and
test used
medium sands, soft silts and
clays without
taking soil samples. The term static implies that
there is no impact on the penetrometer and the device is pushed into the soil by an external
thrust. In this test the tip resistance of a steel cone of 35.7ram end diameter with a projected
2
area of 1(7100 mm and 60 \302\260 point angle is measured while advancing 80 mm into the soil at the
test level at a constant rate of 10 mrn/s to 20 mm/s. The force is applied through a thrust ma-
chine and is transferred to the cone by an assembly of the rods attached to the cone. This as-

sembly is protected by an outer rod (or sleeve).The cone resistance qc is the force required to
push the cone divided by the end area of the cone. The test may be carried out iu 1 ,n to 1.5 m
depth intervals to give the distribution of cone resistance in depth. The data is used to classify
the soil and estimate the allowable bearing pressure and elastic properties via empirical rela-
tionships. The test was originally developed in Netherlands (Dutch Cone Test)and was widely
used in Europe. The developed versions of the cone, which are capable of measuring physical
properties of soils including pore pressure, has gained favour in the USA and Canada. A de-
tailed description of the test along with different types of the penetrometer is standardized by
ASTM D-3441, BS 1377(Part 9) and AS 1289.6.5.1. The test procedures coverthe determina-
tion of the components of the penetration resistance, which includes the cone resistance qc and
side frictionfs.The main types of cone penetrometers areas follows:
Mechanical cone penetrometer. The earliest type had a direct connection to a push rod with a
diameter slightly less than that of the cone. The improved version has hollow outer and solid
inner rods. The inner rod is pushed first and the required force is measured. Next the outer rod
is pushed (with reading of the force) until it rests on the cone. The thrusting is continued on
the outer rod pushing the cone into the soil and thus providing a third reading.

Mechanical friction conepenetrometer. This penetrometer has similar the

design2to mechani\177
cal cone penetrometer with a friction sleeve of surface area of 10000 mm or 15000 mm
addedat the top of the cone. The mechanism of the penetrometer allows the inner rod to push
the cone engaging the friction sleeve after specified advance, thus measuring the cone resis-
tance qc in the first start and the total cone resistance plus side friction f\177 (per unit area of
sleeve) as the test proceeds.
Electrical cone penetrometer. A typical design has tubular push rods and a force transducer
fitted to the cone, which is connected via a central cable (through the push rods) to a surface
data
acquisition system.
Electrical friction cone penetrometer.In this penetrometer the friction force on the friction
sleeve is measured separately.)))
 Bearing Capacity of Shallow Foundations and Piles 431

The push rods assembly form a rigid-jointed string with screwed flush joints, straight axis,
smooth surface, and constant external diameter. The thrust machine should have a capacity of
40 kN to 200 kN and be capableof forcing the penetrometer at a constant rate. Dependingon
the mass of the thrust machine, additional reaction equipment may be required.Test intervals

depend on the type of the project and are 150mm to 200 ram. In a pavement projecta shorter
depth intervals may be used.
The test report include calculation of cone resistance qc at any depth with taking account
the applied force and the weights of the push rods (in mechanical penerometers) and calcula-
tion of side friction fs. These are used to calculate the friction ratio defined by:

Fr = fs in percent)
\303\227i00 (Fr (10.67)
qc
which is used classification and determination
for soil of soil sensitivity. The soil sensitivity St
is defined as of the undisturbed strength
the ratio to the remoulded strength (Section 4,3.7).
The correlations of CPT results to tl\177e engineering properties of soils have been developed
in three major areas of soil classification, the ultimate bearing capacity (either by estimation of
bearing capacity factors or the internal friction angle) and the Modulus of Elasticity. Figure
10.1l(a) shows a soil classification chart reported by Robertson & Campanella (1983). The
lowest curve representing the upper limit for clay soils is suggestedby Bowles (1996).
The undrained cohesion c, may be estimated by applying Equation 10.48:

= / Nk
c\177, (qc -yD) (10.68)

wt\177ere the term

yD represents the total overburden pressure above the tip of the cone and Nt\177is
analogous to the bearing capacity factor No. The magnitude of Nk depends on the clay plastic-
ity and reduces with the increase of the plasticity index (l_unne & Edie, 1976). It varies from
14-21 for normalty consolidated clay and 24-30 for overconsolidated clay. The above investi-
gation also correlates the cone resistance to the vane shear test (Section 4.3.7) results.One
must be aware that the vane shear test overestimatesthe undrained shear strength cu and the
results must be multiplied by a reduction factor 1 to 0.6 for the range of plasticity index 20%
to 100% (Bjerrum, 1974). Bowles (1996) suggestedthe use of the following formula:

N = (13 + 0.i 1PI) (10.69)

k

where PI is theplasticity index. Further in c, may be achieved by adding

improvements the
measured pore pressure (using electrical cone penetrometer) to q,.. The measured pore
type
pressure must be multiplied by 1 - Z where Z is the ratio of the internal cross-section of the
outer push rod (in the vicinity of the cone) to the end area of the cone (Robertson & Campan-
ella, 1983).In sands Meyerhof (1976) presented a chart relating the internat friction angle q'
\302\260 \302\260
(from 30 to 45 ) to the cone resistance apparently without taking into account the effective
overburden pressure. From the Begemann (1974) a series of equations are given below which
evaluate an equivalent bearing capacity factor in the form of:

q,. -- vb \303\227
p; (10.70))))
432 Soil Mechanics: BasicConcepts and Engineering Applications

qo (MPa)
0 10 20 30 4O 5O

5O

1 O0
10 150
2OO

25O

3OO

350

4OO \\40\302\260

(b)
Fr (%)

(a)

Figure 10. I 1. CPTcorrelations: (a) soil classification, (b) determination of effective internal friction an-

gle (Robertson & Campanetta, 1983).

and thus calculating the 0' from:

V = 103.04tanO\"
b (Caquot equation) (10.71)

Vb
= 1.3Xe(2\"5n-e')tan e'[( I + sin 0')I( l+sin2 0')] (Koppejan equation) (10.72)

= 1.3xe2\177tan
\177'tan
2(45\302\260+ \177'/2) (DeBeer equation) (10.73)
Vt,

Another correlation between \177' and cone resistance of the effective

with taking into account
overburden pressure is shown 10.11(b). In
i,1 Figure of average O'from
practice an evaluation
the available equations atld charts is necessary if the
soil deposit is deep and there is no possi-
bility of obtaining undisturbed samples for laboratory testing. The correlation of CPT to the
elastic properties of soils is discussedearlier in Section 5.4.7 where the Modulus of Elasticity
was expressed in terms of cone resistance via empirical equations.However these equations
are based on limited data as conepenetration test is more recent than the standard penetration
test. With the availability of more data in the future the confidence in using the correspondent
correlations will increase (Powrie, 1997). Based on the investigation by DeBeer & Martens
(1957) a compressibility coefficientsimilar to compression index (Chapter 6) was proposedii\177
the following form:

C - l'5qc
(10.74))))
 Bearing Capacity of Shallow Foundations and Piles 433

The settlement of a layer of finite thickness H is estimated from:

H, ,A\177+p\177,

where A<\177is the increase in the stress at the centre of the tayer due to external loading
vertical
and p'o is the effective overburden pressure at that point. This method is applied by dividing
the soil under the footing into finite layers (say up to the depth of 2B) and summing up the set-
tlement of the layers. The above method overestimates the settlement, thus the methods based
on the Modulus of Elasticity expressed by Equations 5.90 and 5,91 may be used. It is com-
monly accepted that the method proposed by Schmertmann (1970) and Schmertmann et al.
(i978) (Section5.4.6)is more reliable in sands and sandy soils.
The relationship between CPT resistanceqc and SPT number N, has been found to depend

upon the average particle size DS0, and the qc I N ratio (Robertson al., 1983;
et B urland &

Burbidge 1985). The following equation is a reasonable correlation of the reported data by the
above authors that is believed to be reliabledata in use:

03
qc =800N(Ds0) DS0_< 1 mm (10.76)
where DS0is the particle size (in ram) corresponding to 50% finer and is determined from the

particle size distribution curve obtainedin the laboratory and qc is in kPa. Note that the energy
ratio corresponding to the N value is 55%. In general the ratio of qc / N varies from 100 (silt)
to 1000 (sand and gravel).

10.4 AXIAL ULTIMATE BEARING CAPACITY OF PILES

10.4.t Pile types

Piles are slender structural members pushed or poured in soils where the strength and settle-
ment criteria are not met using shallow foundations. The major purpose of piling is to transmit
the loads from the superstrvcture m a hard stratum beneath the soil (end bearing pile) or to/he

surrounding soil (friction pile). Other applications include the cases of uplift and overturning

forces, compaction of thick layers of cohesive or cohesionless materials and settlement control
in
slopes and landslides. Based on the method of installation and formation, the piles are clas-
sified into two categories of displacement and non-displacement piles (Weltman & Little.
1977).Displacement piles include preformed, screwed cast-in-place, and driven cast-in-place
piles. In this group there is no removal of soil during installation. Preformed piles are con-
structed from concrete, steel, composite steel-concrete and timber. Driven cast-in-place piles
include a temporary or permanent liner made from concrete or steel.In non-displacement piles
the soil is removed during installation and the preformed pile is pushed into cavity or more

corrm\177oMy the cavity is filled with concrete. The cavity may be supported by permanent or
temporary liner. Permanent liners are made of concrete or steel, whilst the temporary liners are
provided by stabilizing fluid or by soil using continuous flight auger. A combination of both
installation techniques can also be usedto improve the performance of non-displacement piles.)))
434 Soil Mechanics: Basic Concepts and Engineering Applications

Due to the expensive nature of the pile foundations, appropriate sufficient site investiga-
tions must be carriedout according to the relevant building code. This investigation must pro-
vide strength and physical properties of the soil profile, the potential for heave, slopestability,
damage to the adjacent structures and proposed technique for pile installation.

Timber piles. Timber piles have been used for centuriesand are suitable for modest load and
piling length of about 12 m. Timber piles (round or square) have cross sections from 250 mm
(in diameter) to 500 mm (sometimes more).
Working loads are unlikely
500 kN. Al- to exceed
lowable stressesfor compression and tension must be obtained accordingto the building codes
and in particular the reduction factors for compression on the direction perpendicular to the
grains must be investigated. In the Ultimate Strength Design (USD) the strength reduction fac-
tors of up to 0.8 are applied. In the Working Stress Design (WSD) the allowable compression
in the direction of perpendicular to the grains may be adopted as 1/ 5 to 1 / 3 of the allowable
tensile stress depending on the type of the tree used for the pile. The allowable tensile stress
couldbe as high as 17 MPa, whilst the allowable compression is about 8 MPa. Note the aver-
age value of Modulus of Elasticity is at the range of 8000 MPa to 10000 MPa.
Timber piles are driven with pointed end, which may be protected,by a steel or a cast-iron
shoe. Similarly a driving cap may be installed on driving end to protect the butt from hammer-
ing effects. Generally the code has restrictions on the size of the tip and butt as well as mis-
alignment the length of the pile. Timber
along piles have been used successfully in offshore
structures. If the timber is treated with preservatives and is permanently under the water it will

last more than the usual life spans of any structure; however the wetting and drying cycles will
shorten its useful life to few years or less.

Concrete piles. Concrete piles are considered permanent,

but necessary precautions have to be
taken to protect the concrete organic fromsoils. They comprise precastand cast-in-situ piles.
The precast group are displacement type piles and are driven into the ground with the flat or
tapered point and generally equipped with steel or cast iron shoe. The precast concrete piles
are reinforced and have solid square sections of 250 mm to 600 ram, but hexagonal, and hol-
low sections (octagonal and tubular) are aIso used. Hollow sectionsare commonly used in oft'-
shore structures, which have diameter in the excess of 0.9 m to stand high compressive and
bending loads. The precast piles are usually of jointed type to ease the handling and transporta-
tion, but depending on the length of the piles and their numbers, full-length piles may also be
cast in site. Sometimes the reinforcement is prestressed to reducethe sectional area for the
same load capacity.In this way the compressive stresses generatedduring pile driving are bet-
ter resisted, but the ultimate compression strength decreases. The maximum load is at the
range of 900 kN for solid sections and 8500 kN for prestressed piles. The working load may
be taken as the 1 / 3 of the maximum load. The hollow sections are designed for high loads at
the range of 2000 kN or more. The reinforcement details may be found in the relevant concrete
codes. A concrete cast-in-situ pile is made by excavation of a borehole in the ground and fill-
ing the cavity by concrete. Reinforcement is placed centrally after completion of the borehole
and casing is withdrawn during concreting. Normal diameters are 300 mm to 600 ram, but lar-
ger diameters of up to 2 m can be made.When there is more need to base resistance the cast-)))
 Bearing Capacity of Shallow Foundations and Piles 435

in-situ concrete piles are constructed with enlarged bases termed as bell-bottom piles (Mori &

Inamura, 1995).
1989; Tomilinson,
economically cheaper replacement An for a cast-in-situ
concrete pile is a grouted pile, The cavity made by drilling is filled with sand-cement grout
pumped through the hollow flight auger while withdrawing the auger (Gupte, 1984 and 1989).
The main advantages of the grouted piles are minimal vibration and low-level noise. In large
diameters pumpable concrete is preferable to cement-sand mortar.
Durability of concrete
piles must comply with the durability requirements and exposure
classification of the relevant concrete code. Durability requirements include adoption of
minimum concrete strength and restriction on chemical content. Exposureclassification inves-

tigates the chemical conditions of the soil surrounding the pile and commonly covers the de-
termination of sulphates, chlorides and pH value. The details of the concrete pile characteris-
tics, uses and installation are described by Fleming et al. (1992).
Steelpiles.Steel piles comprise H, I and hollow sections. The hollow sections are of tubular or
box type sections. Box sectionsare normally made by welding two sheet pile sections edge to

edge termed as Larssen double box pile. Although they overcome the problems associated
with the of the slender H sectionsthey occupy
flexibility more volume and produce large dis-
placements. The hollow and tubular sections may be filled with concrete. Steel piles are used
when the soil is free from contamination and the corrosion rate is low. Corrosionoccurs in ma-
rine environment and disturbed soils (e.g. man made fills). Corrosionrate could range from
non-aggressive to mild, to moderate, to severe to very severe. The corrosion allowances are
categorized in the building codes under a common title of exposure classificationand are at
the range of less than 0.01 mm/year for non-aggressive exposure to 0.1- 0.5 mm/year for very
severe conditions. In fresh water the corrosion rate is mild to moderate; in the seawater it can
range from severe to very severe for submerging and tide or splash zones respectively.The ac-
celeration of corrosion could occur at the presence of high levels of sulphates, domesticand
industrial waste. Increase of chlorides in soil and water and decrease of pH accelerates the
corrosion.

10.4.2 Basics characteristics of axially loaded pites and pite test

A pile subjected to a force along its central axis carries load partly by shear stresses gener-
the

ated along the shaft (shaft or skin friction) and partly by normal stresses generated at the base
of the pile (Figure 10.12(a)). The ultimate bearing capacity Pu is the sum of the shaft capacity
Ps and the base capacity Pt,. Thus:
= + Ps
P\177, Pb (10.77)
The allowablecapacity of pile Pa is obtained by applying a suitable factor of safety on the
above parts as:

e,, = t'b/ Fb + e\177./ Y.,. \17710.78)

A single factor of safety F = Fb = Fs commonly within 2 to 4 is applied. Substituting values

for the base and the shaft capacities in Equation 10.77 we obtain:)))
436 Soil Mechanics: Basic Concepts and Engineering Applications

Pu

Axial oad

B Total C
eu

........... Shaft

A/\177)-
L
\177 Base

Pile head displacement

(b)

Figure 10.12. Axially loaded pile: (a) base and shaft resistance, (b) load-settlement behaviour.
L
P, = qbAb +
[.\177r.fdA
=
qbAb + r.sAs (10.79)
where qb is the ultimate bearing capacity of the pile base, Ab is the sectional area of the base,
is the limiting shear stress (skin friction) developed on the pile surface,dA is the element of
the perimeter surface area defined by p x dz (p being the perimeter of the sectional area of the
pile) and \177:sis the average limiting shear stress mobilized on the surface perimeter of the pile.
A typical load-settlement behaviour of a pile under axial loading is shown in Figure
10.12(b). This behaviour is consistent with the field data (Whitaker & Cooke, 1966;Bergdahl
& Hult, 1981; Konrad & Roy, 1987) as welt as with the small-scale laboratory tests (Anag-
nostpoulos & Georgiadis,1993). T he idealized response is comprised of three linear portions
of OA, AB and BC. Initially there is a linear increasein the shaft resistance up to point D with
a settlement at the range of 0.5%'to 2% (or 5 mm to 10 ram) of the pile diameter or width
(Fleming et al., 1992).This means that under a working loading below point A the applied
load is resistedentirely by the shaft and little (or no) base resistance is present. In cohesive
soils the shaft resistance is concentratedat the upper part of the pile as indicated by
D'Appolonia & Romualdi (1963) and that the load-transfer mechanism has a time dependent
behaviour (Frances et al., 1961).As the settlement continues (due to loading) the base capacity
is mobilized with an increase in the rate of the settlement (Line BC). At point C the pile head

displacement is at the range of.5% to 10% (more in bored piles) of the pile diameter (or width)
and the base resistance is fully mobilized. Thus in a working condition the axial load Pa is
shared between shaft and base, however the proportion of each share in terms of the total load
cannot be defined. Some estimation is possible using a load-settlement model (Figure)))
 Bearing Capacity of Shallow Foundations and Piles 437

10.12(b)) based on a pile load test. The limiting state of both base and shaft resistance may not
occur at the same time and therefore the validity of Equation 10.77 may not be fully estab-
lished. The mathematical models basedon the load-settlement transfer curves have been suc-
cessfully verified through closed form equations as well as numerical solutions. The closed
form solutions (Cooke, 1974;Randolph & Wroth, 1978) are convenient for homogenous soils
as well as piles with uniform sections. For layered soil and non-uniform (tapered) piles a nu-
merical analysis is more suitable. The accuracy of the numerical analysis depends on the
choice of the static stress-strain model or a dynamic (pile driving) model. DeNicola &
Randolph (1993) and Ghazavi (1997) applied successfullyan explicit finite difference method
FLAC (Sections7.3.4,9.8.2),to verify a closed form equation and a new driving model re-

spectively. Most of the recent computer programs use Smith's model (Smith, 1960), however
new soil models developed by Goble & Rausche (1986) and Lee et al. (1988) and others have
significant applicatiot\177 in numerically based computer programs.
Pile load test is carried out either by static or dynamic loading and is the most reliable
means of evaluating overall bearing capacity. The test procedure must comply with the rele-
vant code (ASTM D-1143, BS 8004, AS 2159). Static load testing is used to evaluate the pile
performance in the early or later stages of the construction work and to proof-test selected
piles.Dynamic pile testing is used to measure the ultimate bearing capacity and its distribu-
tion, parameters to be used in the pile driveability equations, and to assess pile integrity.
Static load is usually applied to a pile by a jack supported by a reaction frame that consists
of a structural beam supported by anchorage or existing coaxial piles. The distance of the an-
chorages and supporting piles from the testing pile must comply with the restrictions of the
code: usually 3B for an anchorage system and 5B for a pile supporting system. The loading

system must be capable of applying the load at a specified rate and maintaining the load for a
specified duration. Loading times ranges from 5 minutes to 15 minutes and could reachup to 6
hours if the load is the representative of the design load. Unloading cycles may be introduced
in the test program if required.
Dynamic testing may be carried out during pile installation or any time thereafter. The test
is performed by using a hammer that is capable of mobilizing the pile strength requirements.
In the pile integrity test the dynamic loading is provided by sonic impact test (SIT), sonic vi-
bration test (SVT) and sonic logging test (SLT). In SIT the pile head is impacted by a small
harrmaer to generate the stress wave whilst in the SVT an electrodynamic vibrator is used
which generates sinusoidal stress waves. In the SLT the sonic pulses are provided by a trans-
mitter to a receiver to determine the sonic profile. The results of the above tests are calibrated
to assess the structural integrity and physical dimensions of the pile.

10.4.3 Basecapacity of piles

For c', 0\" soil the general form of the ultimate bearing capacity of the pile base is:

qb = c'Ncdc+ rlPoN,ldq (10.80)

where rl represents the effect of the at-rest lateral earth pressure. The Nv term is negligible in

comparison with the other terms. Bowles (1996)suggested the use of Hansen's bearing capac-)))
438 Soil Mechanics: Basic Conceptsand Engineering Apptications

ity equations where the pile length L replaces D in the depth factors and rl = 1. Alternatively,
the values of N'c and N'q (instead of Nc and Nq in Equation I0.80) derived by Vesi6 (1975)
may be used:
N c
#
= - l)cot\302\242' (10.81)
(N'q

' 4sinCe'
\342\200\242 /
3\177..,Ie\303\227pl(\17712_,')tan4),]tan2(45.0o + @ (10.82)
Nq =\177
_\177-sin@ { 12)l,.,30-\177sin\177v)i

The term lrr is called the reduced rigidity index and is defined by:

lrr
= Irl(I + \177VIr) (10.83)

where ev is the volumetric strain of the soil at the vicinity of the pile base at failure and Ir is a
rigidity index according:

lr = G /(c'+ P'otan\177') (10.84)

The term G represents the shear modulus of the soil (Equation 5.7). Note that in undrained
conditions ev = 0. The parameter rl is a function of the coefficient of soil pressure at-rest con-
dition ko in the following form:

rl
=
(I + 2k o)/3 (10.85)
For cu, 0u = 0 soil (undrained conditions) the base capacity is reduced to:
= (10.86)
qb cuNcdc

The term Ncdc may be assumed 9 for most practicalpurposes,which is equal to the limiting
value (derived from Equation
of Nc 10,49) proposed by Skempton (1951) for squareor circu-

lar bases. The Vesit's N'c value for undrained conditions is:

N'e =4(]nlrr +1)/3+(rt+2)/2 (10.87)

For piles (c\"

=
0) Fleming et al. (1992)following
in sands Bolton (1986) proposed a solu-
tion based known values of density
on the index Io (Equation 1.13), the critical friction angle
0'or and the effective overburden pressure P'o. Note that the maximum value of the density in-
dex where the field void ratio becomes equal to the minimum void ratio (densest state) ob-
tained in the laboratory is I. The minimum value of Io is zero, which corresponds to the loos-
est possiblestate in the field. The critical friction angle 0\"er is a constant property of the soil
being independent from the initial void ratio and stress level. This is evident from Equation
4.33 (q' = Mp') where q' equals to (o\"1 - #3) and p' is the effective mean stress (o\"l + 2o-'3) /
3. From the geometry of Figure 4.11 or by taking the first term of Equation 4.17 (as there is no
cohesion in the critical state) the magnitude of the critical friction angle is obtained:

sin = 3M/(6 + M)
\177-r

The appropriate value of internal friction angle \177\"under the pile base is estimated from the fol-
lowing equation:)))
 Bearing Capacity of Shallow Foundations and Piles 439

t)\"
=
\177'r
+ 3IO[5\"4 -
ln(p'/Pa)]
-3 (degrees) (10.88)

where Pa is the atmospheric pressure (= 100 kPa). The basebearing capacity is defined by:
=
qb P'oNq (10.89)

Nq is the bearing capacity factor corresponding the

in whict\177 to internal friction angle \177' mobi-

lized beneath the base (Equations 10.88). The effective mean stress p' in the vicinity of the pile
base is taken as the geometric mean of the base bearing capacity and effective overburden

pressure p'o:

p' = p'o (10.90)

To evaluate Nq process is carried out by first assuming an initial
an iterative value for Nq and
calculating p' from i0.90. The corresponding\177' is then calculated from Equation
Equation
10.88 and, using an appropriate d\177'-Nq relationship, the value of Nq can be obtained. If this
value is not sufficiently ctose to the assumed Nq, iteration is continued until the difference in

Nq
between successive cycles is insignificant. The
O'-Nq relationship used is taken from the
theory of ultimate bearing capacity developed by Berezantzev et al. (1961) using a failure
mechanism different to Terzaghi's, Meyerholf's and Hansen's. Whilst the failure mechanism
under the base is similar to Terzaghi's mechanism shown in Figure 10.2, it extends vertically
from points d and d' upwards until intersection with the ground surface, resulting in higher Nq
values than the those from previously mentioned methods. A graphical presentation of Nq ver-
sus \177\"is shown in Figure 10.13.
Correlationsbetween field values of SPT and CPT and the base capacity are very useful,
but the results must be assessed by the methods mentioned above. Meyerhof (1976) proposed

lOOO

100
Nq

25 30 35 40 45
(degrees)
Figure 10.13. Bearing capacity factor Nq used for piles in sand (Berezantzev et al., 196[).)))
440 Soil Mechanics: Basic Concepts and Engineering Applications

the following equation for piles driven in sands:

=
qb 40N'-Lb/B <400\177 (kPa) (10.91)
where N is the average SPT number in the vicinity of 8B (pilediameter or width) above and
3B below the pile base, a,ad Lt, is the length of pile penetration into the sand layer. For fine
sands andan upper limit of 300 \177is
silts taken. Note that in most cases the Equation
10.91
underestimates the base capacity.It is likely that the energy ratio in this equation
55%, and is
therefore the N value must be corrected based on the specification of the equipment used. For
boredpiles a reduction factor in the range of 50% to 67% must be applied.
The averagevalue ofqc within a zone similar to the SPTzone is assumed to be equal to the
base capacity. Fleming & Thorburn (1983) suggested the use of the following equation:

qt,
=
(qcl + qc2+ 2qc3)/4 (10.92)
where qc\177 average cone resistance over2B below
is the the pile base, qc2 is the minimum cone
resistance over 2B below the pile base and qc3 is the average of the minimum values lower
than qcl over 8B above the pile base.

Example 10.12

Compute the base capacity of a round concrete pile that is 25 m long and has a diameter of
0.35 m. The shear strength parameter of basesoil is:c' = 20kPa, qb' = 26 \302\260.
The water table is 5
m below the ground surface, y (above water table)
= 17 kN/m 3, \"\177.rat= 20 kN/m 3.

Solution:
At,
= 7tx0.352 /4 = 0.0962m 2. Using Equations 10.2 we find Nq
= I 1.85, Nc = 22.25.
= 17.0 \303\227
5.0 + -
(20.0 9.81)x 20.0=
288.8 kPa. Depth factors from Equations I0.28:
Po

de = 1+0.4tan-l(25.0/0.35) = !.623,
t
(25.0/0.35) = 1.479.Thus using Equation
= 1 + 2 tan 26.0\"(1-sin 2
dq
26.0\"\177 tan 10.80:

q\177
= 20.0 \303\227
22.25 x 1.623 + 288.8 x I 1.85\303\227
1.479 = 5783.8 kPa.

Pb
=
qt,Ab
= 5783.8 x 0.0962= 556.4kN.

Example 10.13

Re-work Example 10.12 using Vesie's bearing capacity factors with lrr = 10.

Solution:
From Equations and I0.82 we find:
10.81 --- 13. ! 8, Nc = 24.98.
Nq
k o = i - sin d\177'= 1 - \302\260
sin 26.0 0.562, and from Equation
= 10.85: /3
rl = (14- 2 x 0.562)
= 0.708.

= 20.0\303\22724.98x !.623 + 0.708x 288.8x 13.18\303\2271.479 = 4796.6 kPa,

qb
= 4796.6x0.0962 = 461.4 kN.
Pb

Example 10.14
A square pile of B = 0.3 m and length 18 m is embedded in a sand stratum; water table being 4)))
 Bearing Capacity of Shallow Foundations and Piles 441

tn below the ground surface. The prooertiesof the sand layer are: O'er= 30\302\260,
Y (above water
table) = 16.5kN/m 3, Ysut = 19.2 kNlm 3, ID = 0.5. Compute the base capacity of the pile.
Solution:
p\177
=
16.5\303\2274.0+(19.2-9.81)x14.0 =197.5 k.Pa.

Using Equation 10.90 with Nq = 60: p' = -\177\"\177\303\227

197.5 = 1529.8 kPa. From
Equation i0.88:

= 30.0 \302\260
+ 3\303\227
0.5[5.4
- !n(1529.8/100.0)] - 3.0= 31.0% For this value of 10.13
\177' \177', Figure

yields a new value of 38 for

Nq; trying Nq
= 42:
\302\260
p'= -f\177\303\227197.5 = 1279.9 kPa, \177'= 30.0 +3\303\2270.5[5.4 -ln(1279.9/100.0)] -3.0 = 31.3 \302\260 .

From Figure 10.13 Nq = 42. Thus:

qb
= = 197.5 \303\22742= 8295.0 kPa, = =
Pb 8295\303\2270.3\303\2270.3 746.5 kN.
P'oNq

10.4.4 Shaft capacity of pites

In the conventional analysis called ct method (Skempton, 1959;Tomi/inson, 1977) the average
limiting shear stress \"\177smobilized on the shaft is estimated empirically as a fraction of the

undrained cohesion c, in the following form:

z s = \177cu (10.93)

The magnitude of parameter c\177depends on the type of the soil, its past stress history, type of
the pile and the method of installation and varies from 0.3 to 1 (and more). The generally ac-

cepted trend is higher values for soft clays (low shearstrength) and lower values for stiff clays
(high shear strength). This method has been improved (Fleming et al., 1992)relating \177to the

strength ratio defined by cu I P'o according:

, (t.5 ,
\177= (c u I Po) NC (cu / P'o)-\302\260'5 c, / Po -< 1 (10.94)

, 0.5 ,
\177 =
(c, /Po)uc(Cu/ \177,o)-o.\177 cu/p o >1 (10.95)
where the subscript NC represents the normally consolidated state. Equations 10.94 and 10.95

imply that for normally consolidated soil \177-= 1 and th\177 cohesion mobilized on the shaft is

equal to the undrained shear strength. To apply the above equations in pile design it is neces-

sary to obtain the undrained cohesion and effective overburden pressure profiles along the
depth. Thus the shaft capacity may be expressed by:

= (10.96)
Ps \177 (c\177cuAs)i

where \177i, cui are the average values related to a finite length of a pile in specified depth, and
Asi is the perimeter surface area of the finite length. It is recommended by number of authors
that the shaft resistance to the depth of 3B to 4B from the ground surface to be ignored. Based
on the experimental data for driven piles reportedi\177yRandolph & Murphy (1985) Equations
10.94 and 10.95 may be simplified to the following:)))
442 Soil Mechanics: Basic Concepts and Engineering Applications

c\177= 0.5(c u / <1 (10.97)

p\177,)-0.5 cu / Po

c\177= 0.5(c u / p\177)-0.25 cu I p\177

> 1 (10.98)

An aItemative effective stress analysis suggested by Burland (1973) assumes no effectiveco-

hesion on the pile shaft due to the remoulding effects of pile installation. The ultimate fric-
tional shear stressmobilized at a specific depth on the shaft is:

\"\177s
=
oh tan6
= KPo tan\177 =
\177Po (10.99)

where o\"/, is the effective normal stress horizontally applied by the soil on the piIe, 6\" is the ef-
fective friction angle mobilized on the piIe surface, p'o is the effective vertical stress, and K
represents the lateral earth pressure coefficient. Whilst the value of K may be taken as the at-
rest earth pressure coefficient ko for a bored pile (Fleming et al., 1992), this will underestimate
the mobilized shear stress on a driven pile. For bored piles in heavily consolidated soils a
value of K -= (1 + ko) / 2 is recommended. For driven piles experimental results give a range
for K from 1.5 to 1.9 depending on the type of pile (Mansur & Hunter, 1970). For precast con-
crete and H section steel piles, an average value of 1.5 may be used (Meyerhof, 1976). The
American Petroleum Institute's (API) 1984 guidelines recommend K = 1 for driven concrete
piles, but may be assumed to be less for pipe type piles depending on the end condition (open
or closed). For open-end piles a value of 0.8 is appropriate. A 50% reduction in the K value
may be applied where the axial force is tensile. Based on instrumented pile tests Francescon
(1983) recommended K = 1.5 ko. It was also suggested that is
\1775' mobilized to the soil's effec-
tive internal friction angle obtained from simple shear tests.Bowles (1996) s uggested a range
for 6\" from 0.5 \177' to 0.75 #' and emphasized that the method must be used only for granular
materials. Note that the value of\177\" may change with depth due to the increase in effective ver-
tical stress and this has to be taken into account over the length of the pile.
In piles embedded in soft clay negative skin friction may develop when the consolidation

settlement of the layer exceeds the pile settlement. The resulting force in the pile may be esti-
mated using Equation 10.99 with K = ko where ko is calculated from Equation 8.4. If the con-
soIidation settlement is less than the pile settlement then the mobilized shear stress can be
taken as a linear proportion to the limiting shear stress. The mobilized shearstresson the shaft
is equal to the limiting shear stress calculated from Equation i0.99 multiplied by the relative
displacement (ratio or percentage) between the pile and the soil.
Empirical correlations between \177rsand the side frictionf\177 measured in a cone penetration
test may be used with care because of the radial consolidation associated with large diameter
driven piles. Equating zs with fs is acceptable only for slender piles with small volume dis-
placement. Radial consolidation around targe voIume piles may increase zs up to twice the
value offs. Correlations between \177sand the cone penetration resistance qc arealsoused(Flem-
ing et al., 1992). The magnitude of zs varies within the range qc / 10 (Fleming & Thorburu,
1983) to qc 1 40 (Thorburn & McVicar, i971). An empirical equation suggested by Meyerhof
(1976) for piles driven in fine granular materials relates zs to the average SPT number over the
length of the pile:

= 2N
Zs (kPa) (10.100))))
 Bearing Capacityof Shallow Foundatfons and Piles 443

Reduction factors for steel H piles of 50%, and for bored piles of 50% to 67%, must be ap-
plied.
Example 10.15
In an offshore structure a hollow pile of external diameter of 1.2m and length 30 m is embed-
ded in saturated soil with average Ysa\177
= 20.5 kN/m . The profile of undrained coheston \177s

shown in the second column of the table below. \177aiculate the shaft capacity ignoring the con-
tribution of 3B of the length from the ground surface.

Solution:
The results of computations are tabulated where the total shaft capacity is 19.45 MN. Sample
calculation for depth 15 m -20 m:
p\177
= 17.5(20.5 -9.81) = 187.1kPa, c, / p\177 = 575/187.1 = 3.07, from Equation 10.98:
-0.25 = 0.378, = 0.378\303\227575 = 217.3
\177= 0.5 \303\227(3.07) kPa.
\"\302\242s

As = (7t\303\227 5.0
1,2)\303\227
= 18.85 m 2, x As =
\"\177s
217.3 \303\227
18.85/1000 = 4.0961 MN.

Deplh cu P'o \"\177s As \177 Zs x As

Cu / p'o o;
(m) (kPa) (kPa) (k.Pa) (rn') (MN)

0-5 220 26.7 8.24 0.295 64.9 5.28 0.3427

5-10 490 80.2 6.1i 0.318 155.8 18.85 2.9368
10-15 605 133.6 4.53 0.343 207.5 18.85 3.9113
15-20 575 187.1 3.07 0.378 217.3 18.85 4.0961
20-25 535 240.5 2.22 0.410 219.3 18.85 4.1338
25-30 485 294.0 1.65 0.44l 213.9 18.85 4.0320
Total \177
19.45

10.4.5 Axial bearing capacity of pilesusing dynamic formulae and wave equation

A pile driving formula is

idealized equation an used to estimate the ultimate beating capacity
from the pile's to penetration during
resistance driving. It is derived by equating the external
energy induced by each blow of the hammer to the dissipated energy due to the permanent set-
tlement per blow, including an allowance for energy loss. A part of the settlement is due to the
elastic compression of the pile including the pile cap, cap block and ram assembly, and sur-
rounding soil which will be recovered after each blow. There are numerous formulae in use
including the following modified version of the Engineering News Record (ENR) formula that

has been identified as reliable (Bowles.1996):

Wr +
= --
1.25e,\177 E,, n2Wp
(10.IOD
P\177,
s+C w\177+wp

where W\177is the weight of the ram, Wv is the weight of the pile, n is a coefficient of restitution
(to be selected from the relevant code), Et\177 is the maximum rated energy of the ram (from

manufacturer's catalogue), e\177 is the hamraer efficiency, s is the permanent settlement per)))
444 Soil Mechanics: Basic Concepts and Engineer#\177g Applications

blow, and C is the elastic compression of parts. The calculatedPu may be divided by 6 to yield
the allowable bearing capacity of the pile. The magnitude of hammer efficiency eh depends on
the type of the hammer (single action, double action and drop hammers) and varies from 0.75
to i. The elastic compression C represents the half of the recoverable compression and may be
assumed 2.5 mm (0.i inch) for most practical cases. The restitution parameter n is 0.25 for
timber piles, 0.32 for compact cushion on steel pile and 0.5 for steel-on-steel anvil on either
steel or concrete pile.
The dynamic response of a soil during a pile load test may be formulated more accurately
using the one-dimensional pile-soil model originally suggested by Smith (1960). In this model
the length of the pile is divided into discrete elementsinterconnected by springs. Smith's
original model has been modified (Randolph & Simons, 1986) to include the damping effects
of the surrounding soil. The governing differential equation, called the wave equation, is ob-
tained from force equilibrium:

\1772w 1 \1772w T
-- = (10.102)
2 2
Oz \177t
(E/O)p (AE);

where w is the vertical displacement at depth z, E is the Modulus of Elasticity of the pile, O is
the density of the pile material, A is the cross-sectional area of the pile, t is and T is
the time
the mobilized soil resistance per unit length of pile. This differential equation may be solved
using finite differences (Fleming eta[., i992) or be transformed to a finite element form

(Bowtes, 1996)to yield Pu versus settlement as the load test proceeds.

Example 10.16
An pile (HP 310) has the following
H section properties: length of the pile = 12 m, weight per
meter length = 93 kg. The properties of the hammer are: Wr 67 kN, the equivalent stroke =
-\177

0.39m. The pile has penetrated 300 mm in 20 blows. Calculate allowable bearing capacity of
the pile using modified ENR formula. The weight of the cap assembly is 4.5 kN.

Solution:

Ej,
=W
r = 67.0\303\2270.39
\303\227h = 26.13 kN.m. = (93.0\303\227i2.0\303\2279.81/t000) + 4.5 = 15.45 kN.
Wp
=0.015
s = (300.0/i000)/20 m. Take =
e\177, 0.85 and n = 0.5. Thus:

i'25ehEh Wr \177n2Wp i'25x0\"85x26'i3x67'0+0\"52x15\"45 i363.5kN.

s+C W
r +Wp 0.015+0.0025 67.0+15.45

10.4.6 Pile settlement

Settlement of a pile due to shear stresses developed on the shaft is calculated using two simpli-
fying assumptions with regard to the dis*ribution of the shear stress and *he definition of shear
strain. The shear stress at a distance r from the pile axis is assumed to be proportional to the

average shear stress mobilized on the shaft (z\177-\177,):

\"\177
='\177smro/ r (10.103))))
 Bearing Capacity of Shallow Foundations and Piles 445

where ro is the radius of the pile section (or equivalent radius). The shear strain y is approxi-
mated by:

dw
= s - zs,nro
\177t (10.104)
dr rG
where Ws is the settlement of the pile shaft, and G is the shear modulus of the soil. Integrating
Equation 10.104 between r = ro and r = rm (where the settlement becomesinsignificant):

= zs\"r\302\260
w
s ln(r m / ro) (10.105)
G
The force resisted by the shaft equals its perimeter area multiplied by the average shear stress
mobilized on the shaft:

= (10.106)
Ps(working) 2r\177ro L\177o

The settlement under base

the of the pile w\177,is calculated from Equation 5.54 (Section5,4,2).
Alternatively, the base settlement may be calculated using the Mindlin stress distribution dis-
cussed in Section 5.3.1 l. After dividing the soil under the base of the pile into a finite number
of layers (preferably of equal thickness), the vertical stress is computed using Equations 5.50
to 5.53 or Tables 5.2(a)to 5.2(c). The settlement of the base is approximately:

= (10.107)
w/, \177(@/)i

where f is the thickness of finite layer, o\",, is the increase

the in the effective vertical stress at
the centre of the layer due to loading, and E\177 is the Modulus of Elasticity of the soil layer.
Substituting o\"z from Equation 5.50 into Equation 10.107:

If = t
f\177 2
=
13
..... ln, and f] + 12 + 1\177+..- + t n -- L (length of the pile):

i=tl
= = / (10.108)
w\177 (P\177/CE,)y\177(\177q/n)\177 (P\177 Cg.\177.)\177q(\177,,\177r\177,\177)
i 1

which simplifies the use of Tables5.2(a)to 5.2(c),

For stiff piles the total stiffness of the pile-soil system Kp is the sum of the shaft and base
stiffness:

- Pa _ \177(working) Pb(working)
\177-
Kp (10.109)
la;
a W Wb
s

where wa is the total settlement, and Pa is the working load applied to the pile (or allowable
capacity defined by Equation 10.78). With known working values of the shaft and base capaci-
ties, the total settlement of the pile can be obtained using Equation 10.109. For piles that ex-
hibit significant shortening under the working load, a characteristic solution reported by Flem-)))
446 Soil Mechanics: Basic Concepts and Engineering Applications

ing et al. (1992) definesthe ratio Pb(working) / Pa in terms elastic properties of the soil
of the
and the pile. Further information is available from Poulos (1982)and Poulos & Davis (1980).
In order to obtain an estimate of the elastic shortening of the pile, it may be assumed that the
pile forcereduceslinearly from Pa at the top to Pal 2 at the base, which results in an elastic
shortening of:

we = 0.75 Pa LI(A E) ( 10.11 O)

p

Example 10.17
An end bearing pile of 0.4 m diameter 18 m length carries an axial load of 1500 kN. Cal-
and
culate the settlement of the pile by the underlying strata (down to the depth of 18 m
dividing
from the base) into 5 layers of equaI thickness, g = Es
= 28 MPa. 0.3,
Solution:
Thickness of each layer is 18.0 / 5 = 3.6m. Values of m = zI L for r = 0 and the corresponding
lq values (From Table5.2(a))are tabuIated from which the average lq in the zone L to 2L is:
[q(average)
= 23.1891 / 5 = 4.6378,From Equation I0,

wt,
= 0.0138
= [1500.0/(18.0x 28.0x1000)]4,6378 m = 13.8 ram.

z (rn) = z /L t,7
Layer m

1 19,8 1.1 19.3926

2 23.4 1.3 2,2222
3 27.0 1.5 0.8377
4 30.6 1.7 0.4500
5 34.2 1,9 0.2866
Total: 23.189I

10.5PILE GROUPS

10.5.1 Force distribution using the rigid cap concept

For the Ioad distribution through a rigid pile cap Equation 5,95 (Section 5.5.1) may be used in
the foIlowing form:

My Mx p
e, = si(--xi yi
+-\177)
\1771o,111)
ly --i-\177

where P is the total vertical load, Mx and M s are the moments about the x and y axes (Figure
10.14), xi, Yi are coordinates of the centriod of each pile, Ix and/y are the second moments of
area of the pile group about the x and y axes, Si is the cross- sectionalarea of pile i, S is the to-
tal cross-sectional area of the piIe group (excludes the area of the pile cap) and Pi is the verti-
cal load taker\177 by the pile f. The origin of the coordinate system is at the centroid of the pile
group which may be different from the centriod of the rigid cap,)))
 Bearing Capacity of Shallow Foundations and Piles 447

gid ca

\177 .....................
I\"--.---'\177

b c
(a) (b)
Figure [0.14, Risid ca]] s\177pported by a pile group.

For 7, piles of equal cross- sectional area:

M x p
= My,
>, + )0.112>
i
,y,
7

Whilst the magnitude of Pi calculated from the equations above is independent of the orienta-
tion of the xy coordinate system, it is prudent to orient the system so as to minimize the com-
putational effort required.

Example 10.i8

A pile group of Figure I0.i4 carries 600 k,N vertical force at x =y = 0.7 m. The piles are of
equal diameter of 0.35m and sx = 1.4 m. sy = 1.2 m. Calculate the vertical load at each pile.
Solution:
= M = 600\303\2270.7 = 420 kN.m. Considering the right-hand rule sign convention:
IMx] y
2 =
Mx
= - 420 kN.m, My
= 420 kN.m. ,\177,,x\177
= 4x 1.42 = 7.84 m 2, \302\243
y\177
= 6\303\227
(1.2/2) 2.16 m 2,
From Equation 10.112:

PI
= (420.0/7.84)(-1.4) - (-420.0/2.16)(-0.6)+ 600.0/6 = -91.7 kN.
-
P2 = (420.0/ 7.84)(0.0) (-420.0/2.16)(-0.6) + 600.0 / 6 = - 16.7kN.
= (420.0/7.84)(1,4) - (-420.0/2.16)(-0.6)+ 600.0/6= 58.3kN.
P3

Similarly, P4 = 141.7 kN, P5 = 216.7 kN, and P6 = 291.7 kN.

10.5.2 Uhimate bearing capacity and efficiency of a pile group

In a group of closely spacedpiles,or when the rigid cap rests on the ground surface as well as
the a block failure similar to blockabcd of Figure
piles, lO. 14(a) could occur under the ap-

plied loads. In a cohesionless soil the base bearing capacity of the group may be evaluated
from the general bearing capacity equations for shallow footings by substituting the width of)))
448 Soil Mechanics: Basic Concepts and Engineering Applications

the block for B. The frictional resistance can be obtained using Equation 10.99 with K = ko and
\177'= 9'- In sands the total ultimate bearing capacity of the block will be higher than the sum of
the ultimate bearing capacities of the individual piles. In cohesive soils the base bearing capac-
ity is found from Equation 10.49 (Skempton, 1951) by substituting the depth of the block for
D. On all four sides of the block it is assumed that the whole shear strength of the soil is mobi-
lized,tn the case of clay soils the total block bearing capacity may be lessthan the sum of the
ultimate bearing capacities of individual piles. Dependingon the pile spacing a row type of
block failure may also needto be considered (Figure 10.14(b)).
The ratio of the total block bearing capacity to the sum of the ultimate bearing capacity of
individual piles is called the efficiency of the pile group. Jumikis (1971) defines the efficiency
of pile group, independent of the type of the soil, as:

= 1- { [(n
- l)m +
m(n - 1)]/90ran}tan-I (B /s) (1.113)
Eg
where n in the number of piles in a row, m is the number of rows, B is the diameter of piles
and s is the centre-to-centre spacing 04\" the piles.

10.5.3 Settlement of a pile group and Winkler model

The traditional method the settlement of a pile group uses an equivalent raft at a
of estimating
specified depth with loading. However, the current
a nominated surface method combines the
stiffness of the adjacent piles to provide a reasonable soil-pile model (Butterfield & Douglas,
1981) that can be analysed by numerical methods. The settlement of a pile group may be esti-
mated from Equation 10.107 where the vertical stress o\":. due to loading is calculated at the
centre of each finite layer (on each pile axis or any other point) considering the contributions
of all piles. In order to compute cgz the axial force in each pile must be known. With a rigid
cap the axial force taken by each pile may be estimated from Equation 10.111.
When an elastic (uon-rigid) cap rests on a coaxial pile group, the concept of a Winkler
foundation (Section 5.5.2) may be used to evaluate the load distribution and settlements (Fig-
ure 10.15(a)). If the elastic cap is not in contact with the ground surface then the following

procedure may be can'ied out to determine the pile forces and corresponding settlements:

1. Calculate the vertical toad in each pile 10.i i1 if the piles are of different
using Equation
sections. For piles of equal shape and cross-sectional use Equation 10.112.
area,
2. Using the Mindtin solution calculate the settlement of each pile. Decide on the portions of
load to be taken by the shaft and base. The effectof adjacent piles must be taken into account
as well as the pile compression due to the working load and its own weight. Alternatively, any
elastic method could be used to obtain the total displacement of the pile head. Care should be
taken in the selection of the Modulus of the Elasticity of the soil as in most cases it is a func-
tion of the confining pressure and increases with depth.
3. Calculate the pile stiffness (or the spring constarit) of the pile:
Kp
=
P\177 (pile force from /
step 1) wa (pile settlement from step 2).
4. Analyse the cap as a beam resting on finite number of springs with the K/9 values calculated
in step 3.)))
 Bearing Capacity of Shallow Foundations and Piles 449

\177
Elastic beam
', I\177'

illli
............ K (soil)
.........
K (pile) K (pile)
(a)
Figure 10.15. Winkler model for an elastic beam supported by pile group.

If the cap is assumed to rest on the soil as well as the piles (Figure L0.L5(b)), the K value of
the soil has to be considered by replacing the soil between piles with additional springs.

Example 10.19
A rigid pile cap of 2 m by 2 m is resting on two piles and carries a vertical load of 1600 kN at

its centre point. The piles are of diameter 0.4 m and length of 7 m with centre-to-centre dis-
tance of 1.4 m and are positioned on the centre line of the cap. Compute the maximum prob-
able settlement of the pile cap. Assume skin resistance load mechanism, la
= 0.3, Es = 28 MPa,
= 22000 MPa.
Ep

Solution:
To calculatethe settlement at the midpoint of two piles the 7 m thickness of the soil beneath
the base is divided into 5 layers of thickness 1.4 m and Iq values for n\177(under the pile) = rWL
= 0.0 / 7.0 = 0, P\1772
(midpoint between two piles) = rWL = 0.7 / 7.0 = 0.1, n3 (under the other
pile) = r / L = / 7.0 1.4
= 0.2, mt (centre of
layer l)
= z / L = 7.7 / ? = 1.1,m2 = 1.3, m3 = 1.5,

m4
= 1.7, and
m5 = 1.9are tabulated using Table 5.2(c).

Layer z {m) m = ,2/ L tq (rt = 0.0) tq (rt = O. l ) tq

= 0.2)
{\177'1

I 7.7 1.] 3.0612 1.6433 0.7680

2 9.l 1.3 0.8262 0.7411 0.5639
3 10.5 1.5 0.5189 0.4163 0.3679
4 11.9 1.7 0.3332 0.2739 0.2556
5 13.3 1.9 0.2654 0.1965 0.1887
Total: 5.0049 3.2711 2.1441
Average: 1.0010 0.6542 0.4288

The lq in the zone 7 m to 14 m

average is:

lq (pile) = 1,00 [0 +0.4288= 1.4298.

w
b (pile)
= [(1600.0 / =5.8ram.
=0.0058m
2)/(7.0 x 28.0x 1000)11.4298
(middle) = 0.6542x 2 = 1.3084.
lq
w
b (middle)
= [(i 600.0/2)/(7.0 x 28.0x 1000)]1.3084
= 0.0053m = 5.3 ram.)))
450 Soil Mechanics: Basic Concepts and Engineering Applications

If the cap becomes in contact with the ground surface the settlement is calculated from Equa-
tion 5.54. The influence factor for square rigid footing is calculated using Equations 5,55 and
5,58 and is: ls = 0.88.Thus:
S e = (1600.0/2.0\303\227 - = 0.0229 m = 22.9 ram,
2.0)(2.0)(1 0,32) \303\2270.88/28000.0

The elastic shortening of the piles are estimated from Equation 10.110:

w e =O.75PaL/(AE) = = 0.0015 m = 1.5 ram,

p 0.75\303\227800.0\303\2277.0/[(n;\303\2270.42/4)22\303\227I06 ]

The maximum probable settlement = 5.8 + 22,9+ 1.5= 30.2 = 30 ram.

10.6 PROBLEMS

10.1 A square footing of 1 is located at a m surface, soil

depth31.5 The.
properties are: c' = 0, 9 = 40 \302\260,
,m the\177round\177,
= 16.7 kN/m , \177(sat\177-
below?
\1770 kN/m . Using bear-
y Terzaghl's
ing capacity factors caicuIate the ultimate bearing capacity (a) the water table is well be-
low the foundation level, (b) the water table is at tile ground surface.

Answers: 2706 kPa, 1651 kPa

10.2 Re-work Problem i0.1 using Meyerhof's bearing capacity equations.

Answers: 4607 \"kPa, 2811 kPa

10.3 Re-work problem 10.1using Hansen's bearing capacity equations.

Answers: 3596 kPa, 2194 kPa
10,4 Re-work problem 10.1 using Vesi6's bearing capacity equations.
Answers: 4127 kPa, 2518 kPa

10,5 Using Hansen's method calculate the base capacity of a square pile of 0.4 m width and
i0 m length for the following two cases: (a) the water table is well below the pile base
with y = 16.7 kN/m 3, (b) the water table is at the ground surface with \"\177at= 20 kN/m 3.
Ignore the Ny term, c' = 0, 9\" = 40o.

Answers: 2278 kPa, 1390\"kPa

10.6 Re-work Problem 10.5 using Vesid's method with lrr \177-
20,

Answers: 1178 kPa, 719 kPa

10.7 A pile of length is embedded in a sand layer with the followin\177\177properties:
15 m
, o 3 3 . \342\200\242
ID = 0.75, qb cr = = 17 kN/m ,
33 , y =
Y.\177at 20 kN/m . Calculate the base bearing capacity
if: (a) there is \177o water in the vicinity of the pile base, (b) the water table is at the ground
surface. Use the iteration based method of Fleming et at. (1992).
Answers: =19.IMPa, [3,3MPa

10.8 Estimate the ultimate pile capacity of a 30 m concrete pile with 0.4 m diameter in an off-

shore structure where the submerged unit weight is 8,3 kN/m 3. The profile of undrained

shear strength, which changes linearly between the measured points, is:)))
 Bearing Capacity of Shallow Foundations and Piles 451

Depth (m) 0 6 18 24 30
cu (kPa) 200 440 440 220 220

Answer: 5072 kN

10.9 It is required the length of a frictional

to estimate concrete pile of 0.4 m diameter em-
beddedin The ultimate bearing
cohesionless soil.
capacit\177
of the pile is 1850 kN, The
soil is comprised of 10 m fine sand with I = 16.5kN/mJand \177'
= 30 \302\260
underlain by a
3
course sand ofy = 18.8kN/m =
and \302\242\177\"
36 \302\260.
Assume K = 1.5 for both layers.
Answer: 13.5 m
10,10 For the loading system of Example 10.18 a pile group 3 rows with
of 3 piles in each
row is designed. Calculate the vertical force in each pile. Sx= 1.4 m, sy = 1.2 m.
Answers: Left to right: lower row (- 41.7, 8.3, 58.3),middle row (16.7, 66.7, 116.7),
upper row (75.0, 125.0, 175,0), all in kN.

10.11 A pile group 3 coaxial piles all having

has equal length of I0 m and are spaced equally
at 1.5 m. The central pile carries600 kN; the side piles each has 450 kN. Calculate the
= 0.3, Es = 8
settlement of the central pile assuming skin resistance load mechanism, pt
MPa.
Answer: 10 mm.

10.7 REFERENCES

Anagnostpoulos, C. & Georgiadis, M. 1993. Interaction of a\177ial and lateral pile responses. Journal

SMFED, ASCE I 19(4): 793-798.

API, 1984.APf recommended practice for ptanning, designing and constructit\177g fixed offshore platforms,
\177
15 edition. API RP2A page 115.American Petroleum Institute.
Begemann, H. 1974.General report: central and western Europe. Eropean syrup, on penetration testittg..
2.t: 29-39.Stockholm.
Berezantzev, V.C., Khristofornv, V. & Golubkov, V. 1961. Load bearing capacity and deformation of
th
piled foundations. Proc. 5 intern, conf SMFE 2:1 I-15.
\177j\177
Bergdahl, U. & Hult, G. 1981.Load tests on friction piles in clay. Proc. I0 intern, conf SMFE 2: 625-
630.Stockholm.
Bolton,M.D, 1986. The strength and dilatancy of sands\342\200\242
Geotechnique 36(l): 65-78.
\342\200\242 \342\200\242 , th .. .
Bowles, J.E\303\267
1996. Foundatron analys\177s and design. 5 edmon. New York: Mc Graw-Hfll.
Burland, J.B. 1973.Shaft friction piles in clay - a simple fundamental approach. Ground engineering
6(3):30-42.
Burland, JB. & Burbidge M.C. 1985.Settlements of foundations on sands and Proc.,
gravel\342\200\242 Pan 1, 78:
1325-1381. UK: of Civil Engineers.
Institution
Butterfield, R, & Douglas, R.A. 1981.Flexibility coefficients for the design of piles and pile groups.
C1RIA technical note 108\177

Cooke, R.W. 1974. Settlement of friction pile foundations. Proc. conf on tall buildings: 7-19. Kuala

Lumpur.)))
452 Soil Mechanics: Basic Conceptsand Engineering Applications

D'Appolonia, E. & Romualdi, J.P. I963. Load transfer in end bearing steel H piles. Journal SMFED,
ASCE 89(SM2): 1-26.
DeBeer,E.E. & Martens A. 1957. Method of computation of an upper Iimit for the influence of

heterogeneity of sand layers on the settlement of badges. Proc.4 mtern, conf SMFE 1:275-281.
London: Butterworths.
DeNico]a, D.A. & Randolph, M.F. 1993.Tensileand compression shaft capacity of piles in sand. Jour-
nal SMFED,ASCE 119(I2):I952-1973.
Fleming, W.G.K. & Thorburn, S., 1983.Recent piling advances: state of the art report. Proc. conf. on
advances in piling and ground treatment for foundations. London: Institute of Civil Engineers.
net
Fleming, W.G.K., Weltman, A.G., Randolph M.F. & Elson, W.K. 1992. Piling Engineering. 2 edition.
New York: Blackie Academic & Professional.
Francescon, M. I983. Model pile tests in clay: stresses and displacements due to installation and \177xial

loading. Ph.D. Thesis, University of Cambridge.

Gh\177avi, M. 1997. Static and dynamic analysis of piled foundations. Ph.D.Thesis,University of Queen-
sland, Australia.
Gibbs H.J. & Holtz W.G. 1957. Reseach on determining the density of sands by spoon penetration
'. th.\"
tesung. Proc. 4 tntern, conf SMFE 1:35-39.
Goble,G.G.& Rausche, F. 1986. WEAP86 program documentation. FHWA, Washington D.C.
Gupte, A.A. 1984. Construction, design and appIication of post tensioned auger cast soil anchors. Int.
syrnp, of prestressed rock and soil anchors. Post-Tensioning Institute, Des Ptaines, IL, USA.
Gupte, A.A. 1989. Design, construction and applications of auger piling system for earth retention and
underpinning of structures. In J. Burland & J. Mitchell (eds.), Piling and deep foundations I: 63-72.
Rotterdam: A.A. Balkema.
Hansen, J.B. 196[. A general formula for bearing capacity+ Danish geotechnicat institute. Cpenhagen,
Denmark. Bulletin (II): 38-46.
Hansen, LB, 1970. A revised and extended formula for bearing capacity. Danish geotechnical institute.
Cpenhagen, Denmark. Bulletin (28): 5-I I.
,lumikis, A.R. 1971. Foundation engineering. Scranton, Pa: lntext Educational Publishers.
Konrad, J.M. & Roy, M. 1987.Bearing capacity of friction piles in marine clay. Geotechnique 37(2):
163-175.
Lee,S.L.,Chow, Y.K. Karunarame, G.P. & Wong, K.Y. 1988. Rational wave equation model for piIe
driving analysis. Journal SMFED, ASCE 114(3):306-325.
Liao, S.S, & Whitman, R.V. 1986.Overburden correction factors for sand. Journal SMFIz'L), ASCE
112(3): 373-377.
Lunne, T. & Edie, O. 1976. Correlations between coneresistance and vane shear strength in some Scan-
dinavian soft to medium stiff clays. Canadian geotechnicatjournal I3(4): 430-44I.
MacDonald, D.H. & Skempton, A.W. 1955. A survey of comparisons between calcuIated and observed
settlements of structures on clay. Conf. on correlation of calculated and observed stresses and dis-
placements: 318-337.London: Institute of Civil Engineers.
Mansur, C.1. & Hunter, A.H. I970. Pile tests-Arkansas river project. Journal SMFD, ASCE 96(SM5):
1545-I582.
Meyerhof, G.G. 1951. The ultimate hearing capacity of foundations. Geotechnique 2: 301-332.
Meyerhof, G.G. 1953. The bearing capacity of foundations under eccentric and inclined loads. Proc. 3
intern, conf. SMFE I: 440-445. Zurich.

Meyerhof, G.G. 1956. Penetration tests and bearing capacity of cohesionless soils. Journal SMFED,
ASCE 82(SM 1): I-I9.)))
 Bearing Capacity of Shallow Foundations and Piles 453

Meyerhof, G.G. 1963,Some recent research on the bearing capacity of foundations, Canadian geotech-
nicaljournat 1(I): t6-26,
Meyerhof, G.G. 1965.Shallow foundations. Journal SMFED, ASCE 91(SM2):
sr
21-31.
Meyerhof, G.G. 1974. General report: outside Europe, Proc. I European symp, on penetration testing
2.1: 40-48.Stockholm.
Meyerhof, G.G. 1976. [3earing capacity and settlement of pile foundations. Journal SMFED, ASCE
91(GT3): 195-228.
Moil, H & inmnura, T. 1989,Construction of cast-in-place piles with enlarged bases. In I, Burland & J.
Mitchell (eds.), Piling and deep foundations 1 : 85-92, Rotterdam: A.A. Balkema.
Peck, R.B, Hanson, W.E. & Thomburn, T.H. 1974.Foundation engineering, New York: John Wiley &
sons.
Powile, W, 1997.Soilmmechanics-concepts and applications. London: E & FN Spon.
Ramiah, B.K. & Chickanagappa, L.S. 1982.Handbook of soil mechanics and foundation engineering.
Rotterdam: Balkema.
Poulos, H.G. 1982.The influence of shaft length on pile load capacity in clays, Geotechnique 32(2):145-
148.
Poulos, H.G. & Davis, E.H, 1980.Pilefoundation analysis a\177utdesign. John Wiley & Sons.
\177
Randolph, M.F. & Murphy, B.S. 1985. Shaft capacity of driven piles in clay. Proc. 17 offshre technol-
ogy conf: 371-378.
Randolph, M.F. & Simons, H.A. 1986.An improved soil model for one-dimensional pile driving analy-
rd
sis. Proc. 3 intern, conf on numerical methods in offshore piling 3-15. Nantes, France.
Randolph, M.F. & Wroth, C.P. 1978,Analysis of deformation of vertically loaded piles. Jourr]af SMFED,
ASCE 104(GT12): 1465-1488,
Robertson, P.K. & Camanella, R.G. 1983.Interpretation of cone penetration tests. Canadian geotechnical
journal 20(4): 718-745.
Robertson. P.K\303\267,
Campanella, & Wightman
R,G\303\267 A, 1983. SPT-CPTcorrelations. Journal SMFED.
ASCE 109(1I): 1449-1459.
Schmertmann, J.H. 1970. Static cone to compute static settlement over sand. Journal SMFED, ASCE
96(SM3): I011-1043,
Schmertmann, J.H. 1975. Measurement of in-situ shear strength. Proc. conf on in-sire measurement of
soilproperties 2: 57-138.New York: ASCE.
Schmertmann, J.H., Hartman, J.P, & Brown, P.R. 1978.Improved strain influence factor diagrams. Jour-
naf SMFED, ASCE 104(8): 1131-1135.
Seed,H,B\303\267,
Tokirnatsu, K.. Harder, L.F. & Chung, R.M. 1984. The influence of SPT procedures in soil

liquefaction evaluations. Journal SMFED, ASCE 1 | 1(12): 1425-1445.

Skempton. A.W. 1951.The bearing capacity of clays. Proc. building research congress 1:180-189. Lon-
don, UK.

Skempton. A.W. 1959. Castin-situ bored piles in London Clay. Geotechnique 9: 153-173.
Skempton, A,W. 1986. Standard penetration test procedures and the effects in sands of overburden pres-
sure, relative density, particle size, aging and overconsolidation, Georechnique 36(3): 425-447.
Smith, E.A,L. 1960. Pile driving analysis by the wave equation. Journal SMFED, ASCE 86(SM4):35-
61.
Stroud. M.A. 1989. The standard penetration test: its application and interpretation. Penetration testing
in the UK 29-49. London: Thomas Telford.
Te\177aghi, K, 1943. Theoretical soil mechanics. New York: John Wiley & Sons.
Terzaghi, K. & Peck, R. B. 1967.Soil mechanics in engh\177eering practice. New York: John Wiley.)))
454 Soil Mechanics: Basic Concepts and Engineering Applications

r\177l
Terzaghi, K., Peck, R. B., & Mesri G. 1996. Soil mechanics in engineering practice. 3 edition. New
York: John Wiley & Sons.
Thorburn, S. & McVicar, R.S.L. 1971. Pile load tests to failure in the Clyde alluvium. Proc. conf on be-
haviour of piles: 1-7, 53-54. London: Institute of Civil Engineers.
th
Tomilinson, M.J. 1977.Pile design and constrnction practice. 4 edition. London: E. & F.N. Spon.
tla
Tomilinson, M.J. 1995. Foundation design and construction. 6 edition. Harlow, Essex: Long;man
Scientific & Technical.
Vesi6, A.S. 1973.Analysis of ultimate loads of shallow foundations. Journal SMFED, ASCE 99(SMI):
45-73.

Vesi\177, A,S, 1975. Principles of pile foundation design. Soil mechanics services.Schoolof Engineering,
Duke University, Durham. (38): page 48.
Walsh, H. 1981. Tolerable settlement of buildings. Journal SMFED, ASCE 107(ET11): 1489-1504.
Weltman, A.J. & Little, J.A. 1977.A review of bearing pile types. DoE/CIRIA piling development group,
report PG I.
Whitaker, T. & Cooke, R.W. 1966.An investigation of the shaft and base resistance of large bored piles
in London clay. Proc. syrup, on farged bored piles:7-49.London: Institution of Civil Engineers.)))
Index

Active earth (or soil) pressure306 Clay minerals 4

Activity 27 Coarse-grained soils 20, 34
Adsorbed water 9 Coefficient of consolidation 226, 232
Air content 14 Coefficient of curvature 21
Allowable bearing pressure 423 Coefficient of permeability 71
Anchored sheet pile 341 Coefficientof secondary compression 216
Anchors: concrete 351; ground 352; pIate Coefficient of soiI pressure at-rest 303
348 Coefficient of uniformity 21

AngIe of internal friction 115 Coefficient of volume compressibility 215

Auger hole test 81 Cohesion 115

Collapse mechanisms 280

Back pressure 124 Compaction: 41; in the field 46; standard
Base capacityof piles 437 test 43
Basic parabola 102 Compatibility 155
Beam(on an elastic foundation) 203 Compression index 215

Bernoulli's (definition of head) 69 Concrete piIes 434

Bishop's method (general slopestability Cone liquid limit 29
analysis) 393 Cone of depression 77
Bishop's simplified method (stability of Cone penetration: resistance 430,442; test
slopes) 375 430
British soil classificationsystem 40 Consistencystates 26

Bulk density (see density of soil) Consolidated-Undrained test (triaxial) 124

Buoyant unit weight 16 Consolidation:analogy 55; average degree
Burland and Burbidge method 191 231; conceptuaI model 63; correction for

construction period 251; degree 227; set-

Cantilever retaining walls 326 tlement 221 ; test 219; three-dimensional
Cantilever sheet piles 336 settlement 253

Capillary rise 61 Constant head test 73

Cation exchange 8 Continuity equation 92
Classification of soils33 Coulomb wedge analysis 314

Clay 2, 11, 20, 22, 35,36,37, 40 Critical hydraulic gradient 71)))

455
456 Soil Mechanics: Basic Concepts and Engineering Applications

CriticaI state: line 140; theory 137 Failure envelope 116

Crystal structure 4 Falling head test 74
Fellenius' method (stability of slopes) 373
Darcy's law 69 Fine-grained soils 20, 26, 34
Degree of saturation I3 Fixed earth support method 332,342
Density: of soil 14;of solids 14; dry 14; Flexiblewall permeameter 75
Density index 15, 45 Flocculated structure 10
Description of a soiI 37 Flow lines 85
Deviator stress 121 Flow nets 85, 88
Differential settlement 424 IZlow rate: 70; design considerations 106
Dilatancy 35 Flow rule 262
Dilation angle 132 Free earth support method 332,341
Direct sheartest 118 Free water 9
Dispersed structure 10
Displacement diagram 281 Generalshear failure 411
Double layer 9, 42 Geological classification 1
Drained:loading 64; test (triaxial) 124 GraveI 2, 15,20,22,34, 35, 36, 37, 39, 40
Dry density-moisture content relationship Gravity retaining walls 326
41 Group symbols 34
Dry strength 35
Dutch cone test 430 Half closed layer 230, 243
Dynamic compaction 48 Hydraulic conductivity 69
Dynamic formulae 443 Hydraulic consolidation apparatus

(oedometer) 75
Earth pressure (in an excavation) 345 Hydraulic gradient 70
Earth pressure at-rest 303 Hydrometer analysis 22
Earth pressure coefficient(s)303, 308, 311 Hvorslev surface 146
Effective angle of internal friction [ 15
Effective cohesion 115 Illite 6
Effective strcss: concept 55; in paltially Immediate settlement 213
saturated soil 66 ImpermeabIe blanket 89
Effective unit weight 58 Internally stabilized systems 325,354
Elasticity 154 Isochrone 228; parabolic243
Electric charge8 Isomorphous substitution 8
Energy ratio 425 Isotropic compression 137
Equilibrium equations 154

Equipotential lines 85 Janbu's method (general slope stabifity

Excess porepressure55, 64 analysis) 394
Externally stabilized systems 323
Kaolinite 5
Factor of safety: retaining walls 325; shal-
Iow foundations 423; slopes in c., \177,,= 0 Laplace equation 93
soil 368 Lateral earth pressure 275,303, 305,308)))
 lndex 457

Limit analysis 280, 292 Plasticity chart 35, 39

Limit equilibrium method 279, 284, 298 Plasticity index 27

Limit load 261 Plasticity theory 261,265

Limit theorems 291 Plastic limit 27

Linear shrinkage 31 Plate loading test 425

Liquidity index 27 Peat7, 37, 40
Liquid limit 27 Permeability of stratified soils 84
Local shear failure 41 I Permeability tests: in-situ 78; laboratory
Log time method 233 73
Lowerbound solution 263 Phreatic surface 101
Piezomctcr test 82
Mass-volume relationships 12 Pile: Axial ultimate bearing capacity 433;
Modulus of Elasticity 153, 155; (determi- groups 446; settlement 444, group scttle-
nation) 197, ! 98 ment 448; types 433
Modulus of subgrade reaction202 Pneumatic displacement test 82
Mohr'scircleof stresses I 11, 110 Poisson's ratio 153, 155
Mohr-Coulomb (failure criterion) 114, 262 Pore air prcssure 66
Moisture content 14 Poorly graded soil 21
Montmorillonite 6 Pore pressure 55
Morgenstern and Price's method (general Pore pressure coefficients 129
slope stability analysis) 396 Pore pressure distribution (sheet piles) 334
Pore pressure ratio 374,380
Negative pore pressure 62 Pore water pressure 66
Negative skin friction 442 Porosity 13
Newmark's influence chart 165 Potential function 94
Normally consolidated: 214, 222; clay Preconsolidation pressure 214, 218
254;state 215, 218 Primary consolidation 215
Numerical solution: consolidation 240; Primary minerals I

sheet pile walls 353; slopes 404; two- Pumping test 78, 80
dimensional flow 96

Quick condition 72

Oedometer 219 Quicksand 72

Open: boundary 230; layer 243
Optimum moisture content 42 Radius of influence 80
Organic matter 7 Rankine theory 303, 305,345,348, 355
Overconsolidated 192, 198,214; clay 216; Rapid drawdown 380
state 222 Recompression 215
Ovcrconsolidation ratio 215,223,253 Reinforced earth retaining walls 355
I?,csidual shear test I 18
Partially saturated soil 66 Residual soils 2
Particle size distributioll 20 Retaining structures 323
Passive earth (or soil) pressure306 Ring shear test 119
Passive resistance 323,331,333, 345 Root time method 232)))
458 Soil Mechanics: Basic Conceptsand Engineering Applications

Rowe type consolidation cell75 Stabili W earth slopes 385

of nailed

Stability earth slopes 385

of reinforced
Sand 2, 11, 15, 17, 20, 22,34, 35,36, 37, Standard penetration test 425
39, 40 State boundary surface: 140; normalized
Sand cone apparatus 19 148
Sand drains 236, 239 Steady flow 92

Schmertmann's method 193 Steel piles 435

Secondaryconsolidation 215 Stokes' taw 23
Sedimentation: 10; test 21 Stream function 94
Seepage:pressure71;through earth dams Stress discontinuity 264

100; velocity 70 Stress distribution (due to): concentrated

Settlement: aitenaative influence factors load 157; uniformly Ioaded circular area
for saturated clays 190; flexibleand rigid 164;uniformly loaded rectangular area
footings 185; Iayer of finite thickness 187; 167; line load of finite length 169; line
the effect of embedment depth 189; (in) load of infinite length 171; uniformly
sands 191; (in) slopes 403 loaded infinite strip 172; Iinearly Ioaded
Shaft capacity of piles441 infinite strip 175; embankment loading of
Shear box 119 infinite length 176; internal loading 178
Shear modulus 156 Stresspath method (test) 127
Shearstrength 111,114, \177i6, 118, 123, Stress-strain (behaviour) 132, 135
13\177, 143,298 Stress-strain (models) 114
Sheep's foot roller 47 Submerged unit weight 16
Sheetpile walls (static analysis) 331
Shrinkage limit 27 Taylor's slope stability chart 372
Shrinkage ratio 31 Timber piles 434
Sieve analysis 22 Timefactor 227
Silt 2, 10, 11, 20,22, 34,35,36, 37,40, 48 Total active thrust 308
Simple shear test 118 Total head 69
Skin friction 435 Total passive thrust 311, 316
Sliding block analysis 359 TotaI stress 55
SIug test (conventional) 81 Toughness 36

Slurry excavation 346 Tracer test 82

Soil fabric 10 Transported (soil) 10
SoiI-nailed earth retaining walls 357 Tresca yield function 262
Soil structure 10 Triaxial compression test 121
Soil suction 62, 66 Tunnels: heading 296; shallow circular
Specificgravity of solids 14 287, 297; shallow square 288,297
Specific volume 13
Spencer'smethod (general slope stability Ultimate bearing capacity of shallow
analysis) 398 foundations: Hansen's method 417; Mey-
Split barrel sampler 426 erN3fs method 415; Skempton's method
Stability number 369 422; Terzaghi's method 4[3; Vesid's
StabiIity of infinitely long earth slopes 381 method 420)))
 Index 459

Ultimate bearing capacity of strip footing

(limit theorems) 268, 271

Unconfined compression test i26
Unconsolidated-Undrained test (triaxial)
124
Undrained: loading 65; shear strength 143,
197
Unified Soil CIassification System 34, 36
Unit weight 16
Upper bound solution 280

Vane shear test 131

Velocity diagram 281
Velocity discontinuity 280
Vertical cut 273,283
Vertical drains 236
Vibrating plates 48
Vibratory rollers 47
Virgin compression Iine 214,215
Void ratio 13
Volumetric shrinkage 31

Wall adhesion 317

Wail friction 303,316
Water content 14
Water table 56
Wave equation 443
Wedge mechanism: reinforced slopes 389;
unreinforced slopes 400, 401
Welt graded soil 21
Well pumping test (see pumping test)
Wet density (see density of soil)
Winkler model 202.204
Work hardening 261
Work softening 262

X-ray diffraction method 11

YieId function 261

Zero air curve 15, 43)))