FOUNDATION ENGINEERING

NOTES

Owned By:
LOW KINSON
CIV4249: Foundation Engineering 1.1
Topic 1: Site Investigation and Exploration

Topic 1: SITE INVESTIGATION AND EXPLORATION

TABLE OF CONTENTS

1. GENERAL ................................................................................................... 2
1.1. Objectives ............................................................................................ 2
1.2. Reasons for investigating subsurface .................................................. 2
1.3. Sources of Information......................................................................... 3
2. SITE EXPLORATION OBJECTIVES ........................................................... 4

3. SITE EXPLORATION MAGNITUDE AND EMPHASIS ............................... 4

3.1. General ................................................................................................ 4
3.2. Variables .............................................................................................. 4
3.3. Phases of Exploration .......................................................................... 5
3.4. Economics ........................................................................................... 5
4. SITE EXPLORATION METHODS ............................................................... 6
4.1. Geological Investigations ..................................................................... 6
4.1.1. Surface studies .......................................................................................... 6
4.1.2. Seismic Surveys: refraction (land) or reflection (marine).......................... 6
4.1.3. Resistivity Surveys ..................................................................................... 7
4.2. Trial Pits ............................................................................................... 7
4.3. Boreholes ............................................................................................ 7
4.3.1. Auger holes ................................................................................................ 8
4.3.2. Percussion Borings .................................................................................... 9
4.3.3. Rotary Wash Boring ................................................................................... 9
4.3.4. Rotary Core Drilling (also diamond drilling) ............................................. 10
4.3.5. Borehole Support ..................................................................................... 10
4.3.6. Borehole Logging ..................................................................................... 11
4.4. Insitu Testing ..................................................................................... 11
4.4.1. Standard Penetration Test (SPT) ............................................................ 12
4.4.2. The mechanics of the SPT test................................................................ 12
4.4.3. Cone Penetration Tests (CPT) ................................................................ 16
4.4.4. Pressuremeter Tests (PMT) .................................................................... 21
4.4.5. Dilatometer Tests (DMT) ......................................................................... 25
4.4.6. Vane Shear Tests (VST).......................................................................... 26
4.4.7. Field Permeability Tests .......................................................................... 27
4.4.8. Applicability of Test Methods ................................................................... 27
4.5. Sampling ............................................................................................ 28
4.5.1. "Disturbed" Sample .................................................................................. 28
4.5.2. "Undisturbed" Samples ............................................................................ 29
5. SITE EXPLORATION PLANNING............................................................. 29

6. SITE INVESTIGATION REPORTS ............................................................ 30

7. REFERENCES........................................................................................... 31

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.2
Topic 1: Site Investigation and Exploration

1. GENERAL

1.1. Objectives

The practice of site investigations is undoubtedly one of the most critical aspects of
the geotechnical engineers’ work. The broad objective of site investigation is to
accumulate all data for the site in question so that the proposal being considered
may be satisfactorily carried out. It follows that the site in question must be examined
thoroughly so that subsequent analyses and design procedures are appropriate to
and representative of the proposal being considered. The type of information
required will depend on the proposal as indicated below.

(a) Investigations for new works: access, rights of way, other rights and
covenants, location of services, presence of mine workings likely, effect of
works on adjoining properties, ground conditions, records of performance of
adjacent structures, topographic, hydrologic and climatologic data (dams),
environmental considerations such as land use population etc.

(b) Investigations of failures (and possibly remedial design measures):

observation of mode of failure, past or continuing movements, specialized
investigations such as ground conditions or structural data etc. depending on
the nature of failure. This group could also include investigations of pollution
problems.

(c) Reports on existing works (perhaps on likely adverse effects from new works
or the possibility of enlargement or modifications): all available records
concerning design and performance of original structure, observation of
movements and/or movement records, ground conditions relevant to the
problem.

(d) Investigation of availability of materials (for fills, road bases, concrete

aggregate, dam core filters or fill etc.): largely reconnaissance and geological
data drilling to estimate extent of deposits and possible laboratory tests to
assess properties and/or suitability for treatment [e.g. lime stabilization etc.].

1.2. Reasons for investigating subsurface

(a) To establish site suitability;

(b) To enable effective, efficient and economic design and construction;

(c) To anticipate immediate and/or future problems and apply appropriate

measures;

(d) To give confidence in the design assumptions;

(e) To give insurance against legal actions.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.3
Topic 1: Site Investigation and Exploration

1.3. Sources of Information

(a) Experience and general knowledge of the locality. (Basis for scientific
investigation may give warning that otherwise might be overlooked etc.).

(b) Existing records e.g. geological survey maps and memoirs; ordinance survey;
municipal and public authority plans; meteorological records; borehole records
from adjacent significant areas; aerial photographs.

(c) Personal onsite investigations: inspection of adjoining areas; collection of hand

specimens to correlate with geological maps; topographical features especially
instability, drainage, fault evidence, vegetation. Talking with locals especially
land owners, farmers, local builders etc. Terzaghi maintained that the absolute
minimum site investigation was "to sit and stare" - no doubt he thought as well.

(d) Site Exploration (with field and laboratory testing). This is the investigation of
the subsurface conditions and forms the main part of this course.
Unfortunately its practice is still very much neglected particularly on small
infrastructure projects. Often a site investigation is not ever undertaken. In
these instances problems only become apparent on construction with the
result that delays for remedial measures and/or redesign operations which
increase costs quickly often overtaking by considerable amounts the capital
which would have been required for the site exploration. In other instances,
the site exploration may have been of poor quality with the result that soil
stratification and parameters are not representative of the subsurface
conditions. This situation can lead to two extremes:

 Gross overestimation of suitability of the subsurface soils resulting in

excessive settlements, bearing capacity failure in extreme cases, ground
water problems, shorter pile lengths etc. etc. One very serious aspect of
this extreme can occur if the depth of exploration does not extend to a depth
which is significantly stressed by the proposed construction e.g. settlement
of pile groups.
 Gross underestimation of suitability of the subsurface soils resulting in very
conservative hence costly foundation design.
It follows from the above that some form of site investigation is desirable for almost
every form of infrastructure development. The scope and extent of the exploration
will depend on many variables, but its absolute exclusion can prove to be a very
false economy.

More important it is not just sufficient to undertake any old site exploration as this
can lead to further excessive expenditure whether of the form of remedial measures
for overestimation of site suitability or costly foundations for underestimation of site
suitability. The site exploration should be carefully planned, undertaken and
supervised in relation to the proposed developments and the subsurface conditions
encountered.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.4
Topic 1: Site Investigation and Exploration

2. SITE EXPLORATION OBJECTIVES

The precise objectives of the site exploration will be determined by the nature and
scope of the geotechnical engineers' brief. It should be noted however that any
unusual conditions encountered and not necessarily in the brief should also be
reported if considered relevant.

In general the objectives would be as follows:

 To determine the nature, thickness, dip and variability of the strata to be affected
by the proposed development;
 To determine the physical properties of the strata relevant to the design;
 To determine the ground water level (or levels) variation and chemical
composition;
 To obtain the above data in the most economic and technically accurate way.
This requires flexibility in planning and good knowledge of all available
techniques.

3. SITE EXPLORATION MAGNITUDE AND EMPHASIS

3.1. General

The aim must always be to provide adequate reliable information. If compromise is

necessary (inadequate funds) then obtain reliable information at the expense of
adequacy. Some site exploration is essential in all civil engineering and building
works. The extent depends on information available, relevant variables,
consequences of failure, complexity of foundation, magnitude of loads, geology of
site and ease of interpretation, overall job economics etc.

3.2. Variables

Apart from economics, the major considerations governing the magnitude and
emphasis of the exploration are the types of soil profile and the type of development.

(a) Two soil profile extremes are possible (and of course everything between).

Profile Very variable mixed soil Very uniform soil

Exploration Much field investigation Little field investigation

Characteristics Simple laboratory tests Considerable lab testing

Little undisturbed testing

(b) Two development extremes are possible:

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.5
Topic 1: Site Investigation and Exploration

Development Large areal extent Limited areal extent

Exploration Diffuse Concentrated

Characteristics May be shallow Usually deep

The proportions of disturbed to undisturbed sampling and of field to laboratory

testing will depend on the job and the types of ground encountered.

3.3. Phases of Exploration

Particularly with a major development on a limited area it is sometimes advisable to

carry out the exploratory operations in three phases.

(1) Preliminary Explorations to asses the general suitability of the site for the given
development or to select the most suitable site from alternatives. This phase
and its results will enable a preliminary assessment of soil and ground water
characteristics so that subsequent phases of exploration can be planned -
location of boreholes, depth, and likely size of boreholes, testing techniques
and equipment necessary. In areas, which are well documented, this phase of
exploration is not always necessary.

(2) General Exploration to select areas of a given site with the most suitable
ground conditions or simply to investigate the ground conditions of a fixed
location. To obtain the required samples and conduct the appropriate testing.
Usually this phase is sufficient for general construction.

(3) Detailed Explorations generally for use in major, heavy or unusual

developments. Special techniques of testing and explorations may be used to
examine specific soil behaviour when conditions require it. This may include
plate bearing tests, specialised sampling and testing techniques, large
diameter samples, trial embankments etc. etc. Usually this phase can be quite
an expensive operation and is therefore only really applicable to multimillion
dollar projects or projects in which the consequences of failure could be
serious.

3.4. Economics

Frequently, a limiting factor on magnitude. However, an adequate site investigation

will generally cost between 0.25 and 1% of the gross project cost under average
conditions. The percentage generally goes down as the total project cost up and
can get as low as 0.1% Bridges, wharves and similar structures require relatively
larger investigations than multistorey buildings. In many cases of domestic
constructions, a personal inspection of the site, hand augering to a small depth or
even inspection of the deeper stump holes, field identification and classification of
the soil types and soil profile would be sufficient. The lack of detailed information is
compensated by a liberal factor of safety. On large jobs, the cost of thorough and
elaborate investigations is usually small compared with the savings realized by
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.6
Topic 1: Site Investigation and Exploration

effective use of the data in both design and construction (e.g. tender process for
below ground work will often be drastically reduced if contractor is confident of
subsurface conditions).

4. SITE EXPLORATION METHODS

Exposure of relevant strata for inspection, sampling or testing may be by trial pits or
shafts, boreholes or audits. On the other hand, geological methods of investigation
may provide considerable data without even disturbing the ground surface. It is
recommended that some form of boring is undertaken for correlation of results, and
positive identification. The choice of methods will depend on:

 type of information and details required;

 availability of equipment;
 costs;
 utilization of existing openings or exposures;
 possibility of future utilization of any exploratory exposures e.g. holes for
ventilation shafts or service ducts;
 access to site and working space;
 anticipated ground conditions;
 weather;
 time;
 human factors

4.1. Geological Investigations

4.1.1. Surface studies

(a) Geological Map (e.g. Australian Geological Survey) should always be the first
source of information. This provides a preliminary indication of what to look for
and what to expect; geological succession and general description; contact
zones; faults dips and strikes.

(b) Field Traverse: On major projects, geologists will be available. The engineer
must understand enough to request a geologist when necessary and to
appreciate the results. On small jobs he must be his own geologist.

(c) Aerial Photography: Photogeology is a rapidly expanding expert field. A great

deal of information can be obtained from such photographs particularly from
stereo pairs e.g. faults and other major zones of weakness, areas of intense
folding and faulting, stream alluvium, alluvial fans and plains, glacial deposits,
sand dunes, major rock outcrops, areas of fill etc. Infrared photography can
also provide useful information on areas where water is present.

4.1.2. Seismic Surveys: refraction (land) or reflection (marine)

Seismic survey methods are based on placing a string of geophones along the
ground surface, and measuring the time taken for a sonic event (such as a sledge
hammer hitting a metal plate, or a small explosive charge) to be recorded at each of
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.7
Topic 1: Site Investigation and Exploration

the geophones. The sonic wave first reaching a geophone may have travelled
through the ground directly below the surface, or may have travelled to a deeper
more competent layer with higher sonic wavespeed and been refracted up to the
geophone. Knowing the distance between each geophone and the event allows
geologists to predict a model of the subsurface profile based on sonic velocities.
Some features of the method are:

 The most accurate and sophisticated of geophysical techniques;

 Somewhat expensive, but valuable as a first brush method of investigation
to establish the main subsurface features;
 Good for areas where vehicle access is difficult.
 Permits location of bedrock and major stratigraphic layering, faults, dip of
strata etc.;
 Can establish the (small strain) elastic modulus;
 Drillholes are essential for calibration, but may be located on the basis of
the inferred seismic results;
 With complex (or even just slightly complex) stratification, interpretation can
be inaccurate;
 A good supplement to conventional exploration techniques, but never likely
to replace these techniques;

4.1.3. Resistivity Surveys

Cheaper than seismic survey but less accurate. For location of bedrock or faults and
determination of groundwater salinity. Generally better for qualitative rather than
quantitative information.

4.2. Trial Pits

 Cheapest method of exploration to shallow depths. Uses backhoe or may

be dozer;
 Safety is an issue for anyone entering a trial pit greater than 1.5m deep;
 Enables hand cut samples with minimum disturbance;
 Gives a clear picture of stratification and variability;
 for rocks direct evidence of dip fissuring, jointing, weathering in joints and
fissures etc.;
 No use below water table in granular soils;
 gives good indication of quality of fill size of voids etc. in made ground;
 Useful for inspection of old foundations;
 Should always be used for examination of surface soils if there is the
slightest doubt or anomaly raised in the borehole logs.

4.3. Boreholes

Boreholes constitute a critical element of exploration. All interpretation and

subsequent testing depends on the quality of sampling and the adequacy of site
logging. Hence all drilling should be under the FULLTIME control of a field engineer.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.8
Topic 1: Site Investigation and Exploration

4.3.1. Auger holes

[Link]. Hand Auger

Hand augering is used for very simple investigations (e.g. house foundations) or
where access would otherwise be difficult or impossible. The hand auger is simply
a fence-post auger. Different heads are available for different soil conditions.

 Usually a bucket auger 40mm - 200mm diameter;

 Generally to a depth of about 2m depending on diameter;
 Undisturbed sampling possible;
 Almost always in uncased holes (i.e. no equipment required);
 Difficult with granular soils;
 Power augers are also available - these are spiral flight augers with a small
petrol engine mounted above. Penetration is much faster, but generally two
persons are required.

[Link]. Solid Flight Mechanical Auger

These augers are often truck-mounted and penetration to very substantial depths
(many 10’s of metres) is possible by progressive addition of more auger sections.

 With full flight augers, a worm flight brings the soil to the surface. With other
types, the auger is withdrawn regularly to inspect soil.
 Depth from which soil is taken can be reasonably estimated.
 'Undisturbed' sampling possible by withdrawing auger. Disturbance below
bottom of hole is very small.
 Advancing in sands can be difficult due to borehole collapse. Advancing in
sands below the water table can be impossible.
 75mm to 300mm diameter is common for site investigation but up to 1m
sometimes used for inspection bores.

[Link]. Hollow Flight Mechanical Auger

This is a piece of apparatus which is not commonly used in Australia.

 Centre post of auger is hollow tube with plug, which can be removed for
sampling.
 Auger (full flight) need not be withdrawn for sampling. Therefore, no hole
support is required.
 Good access provided to the bottom of hole with no disturbed material able
to fall in and low boring disturbance ahead of auger.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.9
Topic 1: Site Investigation and Exploration

4.3.2. Percussion Borings

Percussion boring is a very rapid method of drilling a hole into the ground. The drill
is progressively advanced into the soil/rock, and not withdrawn until the target depth
has been reached. Very disturbed soil and rock cuttings are typically blasted from
the top of the hole from the air pressure used for the percussion.

 A mechanical means of advancing holes which requires relatively light

equipment;
 Lift and thump boring requiring - this is usually achieved by compressed air,
but could also be achieved with high pressure water;
 Percussion allows penetration through even hard rock, but advancing
through soft soils can be difficult;
 Strength of clays or density of sands may be inferred from rate of
penetration;
 Tools vary from open ended steel tube with cutting edge for clays (clay circle
or 'auger') through open-ended steel tube with cutting edge and flap valve
for sands (sand pump or shell) to various types of 'chisel' for gravels and
rock;
 Deep disturbance ahead of hole undesirable hence not recommended for
general investigation work. 'Undisturbed' samples not possible, but
disturbed samples obtained continuously;
 Only general (not detailed) stratigraphy can be established.

4.3.3. Rotary Wash Boring

Rotary wash boring is possibly the most commonly used method of drilling used for
geotechnical site investigations in Australia. It is also the method which causes least
soil disturbance. Principle features of this method are as follows:

 Allows moderate rate of advancement of holes with reasonably light

equipment;
 Soil is loosened and removed from the borehole by water which flows down
the washpipe and up inside the borehole (or casing if used);
 Cutting bit on the end of the washpipe assists scouring action of water in
loosening soil;
 Different cutting bits are used in clays (tri-cone bit), sands and gravels
(spade bit);
 Use of drilling mud in sands allows the hole to remain uncased. Casing not
too often necessary in clays.
 Wash borings are settled out in a setting tank and can be identified very
disturbed.
 Soil below bottom of hole very little disturbed.
 Undisturbed sampling possible.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.10
Topic 1: Site Investigation and Exploration

4.3.4. Rotary Core Drilling (also diamond drilling)

Rotary core drilling can be used in association with either augered or rotary wash
bored holes. When competent rock is encountered, these techniques are not able
to advance any further because of the crowd (downward thrust) and torque
limitations of the drill rig. Undisturbed sampling of the rock then requires diamond
coring techniques. With diamond coring, it is possible to establish not only the rock
characteristics, but also evaluate the defects such as joints and fault zones.
Effectively unlimited penetration through the rock is possible, although penetration
rates are very slow (typically 0.5 to 6m per hour depending on rock hardness).
Essential features of this technique are as follows:

 Used for very hard materials;

 Double or triple tube core barrels generally used. Inner (sampling) tube
remains stationary while outer barrel and drilling bit rotate;
 Diamond chips on the bit cut an annulus in the rock or hard clay;
 Spring steel core catchers can be used to ensure core is retained;
 Provides sample as well as making borehole;
 Drilling water or mud cools cutting bit brings chips to surface;
 Core barrels of 1.5m and 3m lengths are most common;
 At end of core run, core barrel is returned to the surface, and the core is
extruded from the barrel into split PVC tubes which are wrapped in plastic
tube before being stored in a core box. Cores are generally photographed.

4.3.5. Borehole Support

Boreholes will usually stay open in clays which are stiff to hard. However, there is
a likelihood of borehole collapse in soft clays or granular soils, so some means of
maintaining the hole open is required. The methods used are either steel casing or
drilling mud.

[Link]. Casing

Casing requires the use of steel tubes in a borehole to support the sides. It can be
installed by driving but is more commonly advanced by hydraulic pushing -
remember - it has to come out again! Casing should never be driven ahead of the
base of the borehole indeed it should be kept as far behind as possible.

Where deep boreholes are to be installed the telescoping of casing may be required
in order to overcome installation resistances. The use of casing has the
disadvantage that it must be carried to site. It takes time to install and of course
remove and it involves capital and maintenance costs. However for shallow holes it
is quite efficient and effective.

[Link]. Drilling Fluid

A common alternative to casing is drilling fluid. The original drilling fluid was based
on natural bentonite clay, which can be mixed in powder form to the drilling water to

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.11
Topic 1: Site Investigation and Exploration

create a higher density suspension. Also, the use of synthetic polymer (e.g., partially
hydrolysed polyacrylamide) as drilling fluid has gained significant appreciation. The
support fluid keeps the borehole stable by applying a hydrostatic fluid pressure
(greater than the pore water pressure) on the base and the side wall. The use of
drilling fluid must be evaluated in terms of its environmental effect on the
groundwater regime. Also, not all fluid types (bentonite/polymer) are suitable for all
ground types.

4.3.6. Borehole Logging

The means of advancement of the borehole are very important in ensuring that
samples retrieved or tests performed are representative of insitu conditions.
Another key requirement is that the drilling procedure and result of the sampling and
testing be logged accurately and in detail. All holes should preferably be logged by
a professional engineer. Logging should include:

 Drilling contractor;
 Client;
 Location;
 Borehole number;
 Type of boring;
 Date;
 Position and elevation of bore;
 Size of casing;
 Rate of progress recovery;
 Water table/s and variations with time;
 Upper boundary of each new stratum (by sampling or examination of wash);
 Borehole support used;
 Field classification;
 Depth and full description of all samples (soil and water)
 Results of SPT tests or other insitu tests;
 Results of any quick tests;
 Type of tool used;
 Reasons for delay or changes in technique;
 Any unusual occurrences;

4.4. Insitu Testing

This testing method is valuable for:

 Quick reconnaissance and preliminary design data;

 Situations where sampling is difficult or gives excessive disturbance (e.g
sands);
 Permeability measurements especially in stratified and lensoidal deposits.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.12
Topic 1: Site Investigation and Exploration

4.4.1. Standard Penetration Test (SPT)

The following sections describe the mechanics of the SPT test, and some of the
many methods proposed for correlating the results engineering properties.

The SPT is one of the most common forms of in-situ test, and is regularly used to
determine the properties of granular soils. The technique is very crude but has the
major advantage of being cheap and easy to carry out. Also the results it gives
whilst far from rigorous are generally accepted right around the world.

4.4.2. The mechanics of the SPT test

W a te r p o rt
H ead P in 465m m 60m m

35 mm

50 mm
D r iv in g
D rillin g B a ll S p lit shoe 20m m
ro d v a lv e b a rre l T h re a d s
C o u p lin g

The equipment consists of a 50 mm outside diameter open-ended tube (not just a

tube but a special spilt tube sampler) with all other dimensions standard. This tube
is placed at the bottom of a
borehole and is connected to the
surface by means of a set of drill
rods. The tube is then driven into
the soil at the bottom of the
borehole by means of a hammer
arrangement which is attached to
the top of the drill rods. The
hammer applies a standard force
per blow provided by a mass of
63.5 kg dropped through 760 mm.
A detailed setup of this test is
shown below.

The number of blows required to

drive the tube for 3 successive
distances of 150 mm (450 mm in
total) are counted. These are
generally reported as x/y/z. The
blows for the first 150 mm (x) are
discounted on the basis that the
soil at the base of the borehole
was disturbed. The number of
blows to drive the tube through the
next two lengths of 150 mm (y/z)
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.13
Topic 1: Site Investigation and Exploration

(300 mm in total) are added together to give the "N" value for the SPT. By conducting
a reasonable number of these tests in the granular formation a reasonable average
N can be obtained. If the ground at the site is hard/very dense then SPT refusal
(N>100=R) is often encountered where a solid SPT bit is recommended to avoid
damage to the standard hollow-bit. If a solid SPT is done then the collection of
sample from the test depth cannot be performed.

The measured N-value at a given depth for a given soil can be influenced by several
factors (e.g., hammer efficiency=ratio of actual energy to the sampler to input energy
(Wh), borehole diameter, sampling method, rod length). Amongst all the hammer
efficiency is the most influencial; a lower efficiency hammer would produce higher
N value which will produce unrealistically higher values of relevant soil strength
parameters or vice-versa. To avoid this issue, the measured N blow count is often
standardised against an average hammer energy of 60% and the converted N is
termed as N60. Hammer efficiency values for different types of hammers (e.g., donut,
safety) used in different countries can be found in standard foundation engineering
text books including some other minor corrections stated above. In case of no
information on the hammer efficiency is available then an industry recommeded
value or assumption is to be utilised.

The hammer efficiency is to be corrected for all soil types. The N60 obtained from
this process will further subject to dilatancy and overburden corrections as
discussed below.

[Link].1. Dilantancy correction for fine or silty sands below the water table

Time The reason for this correction is that for a N60

greater than 15 the sand is medium dense or
denser. This means that when the hammer is
pwp
dropped to drive the tube the penetration of
(-ve) the tube causes the sand to dilate. Where
water is present this dilation causes negative
Cleansand
porewater pressures to be developed.
SPTblows

For relatively clean sand these porewater

pressures are fully dissipated by the time the
Time next blow is applied. For sand which has its
pore space partly blocked by silts or clays or
Buildupin is a very fine sand the rate at which these
pwp negative pwp negative porewater pressures dissipate may
(-ve) be very much lower. Therefore, they may not
be fully dissipated by the time the next blow is
Fine sandor applied or so on. This may mean that there is
silty/clayey a build up in negative porewater pressures as
sand
the blows are applied.

Since there has been no change in total stress

a reduction of porewater pressure (even though it is temporary) must lead to a
temporary increase in effective stress leading to a higher blow counts. Since the
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.14
Topic 1: Site Investigation and Exploration

greater the effective stress the greater is the strength of the material being tested.
Therefore, as the strength increase has only been caused temporarily because of
dilation effects, it must be accounted for by some means. Hence the application of
the above correction to reduce the measured N60 value. As this correction is often
needs a judgement on soil permeability, there are different opinions amongst
practicing engineers.

Note that for this correction to apply the sand being tested must

 be below the water table

 have an N60 > 15
 have reduced permeability.

[N60]d=15+0.5*(N60-15) where [N60]d is the corrected N60 for dilatancy.

[Link].[Link] correction for depth

When correlations were originally made between SPT N values and various
engineering properties a major oversight (error?) was made. This oversight involved
N values obtained at shallow depths. It was found that predictions of soil strength
were unrealistically low and compressibility high. Therefore in order to correct this,
a correction factor which depended on the effective overburden pressure at the
depth at which the test was carried out is introduced. Many relationships were
developed and can be found in the text books. However, one common form is that
if the SPT is conducted at depths having effective vertical stress < pa=100 kPa then
the N value is to be amplified and a reduction is applied for effective stress > 100
kPa. A typical correction factor (CN) as proposed by Liao and Whitmann (1986) is
given below. One may wish to use some other relationships and obtain an average
value as the final correction factor. The corrected N60 is term as (N1)60=CN*N60.
.

𝐶 where, ’o=effective stress, pa=atmospheric pressure (=100 kPa)

[Link]. Correlations - Relative Density, Friction Angle, Modulus of Elasticity

Relative density (Dr%): The packing of granular soils can be described by its
relative density. The higher the relative density of the soil the higher is the density.
It has a scale from 0 to 100 with the lower value represent a loose packing (i.e. 0-
15=Very loose soil) and the higher value refers to very dense packing (i.e.,
>85=Very dense soil). The packing has direct relationships with its strength and
deformation properties (e.g., internal friction angle (’), Youngs modulus (E)).
Therefore, estimation of Dr% in terms of N is essential for granular soils as
undistustubed samples cannot be collected for laboratory tests. Any correlations
between Dr% and N must also include the depth or effective stress to reflect on the
real field condition. Some simplified correlations are given below and the readers

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.15
Topic 1: Site Investigation and Exploration

are advised to learn many other complex form of the correclations from the text
books.
.
Meyerhof (1957): 𝐷 Where, Dr not in % and applicable
.
to clean and medium to fine sands.

Skempton (1986) suggested an approximate correlation between the corrected N

and Dr(%) for sands with Dr(%)>35%.

Skempton (1986): 60 Where, 𝑁 should be multiplied by 0.92 for

coarse sands and 1.08 for fine sands.

Friction angle (’): The peak friction angle for granular soils can be approximately
correlated with the N value. Some approximate corrections are given below.

Peck et al. (1974): 𝜙 𝑑𝑒𝑔 27.1 0.3 𝑁 0.00054 ∗ 𝑁

.

Schmertmann (1975): 𝜙 𝑡𝑎𝑛

. .

Hatanaka & Uchida (1996): 𝜙 15.4 𝑁 20

Drained Young’s Modulus (E’): Empirical correlations between the N and drained Young’s
modulus for sands must be treated with caution. Nevertheless, these correlations can be
used as first-order estimation for E’. Kulhawy & Mayne (1990) developed the following
correlations where, pa=atmospheric pressure=100 kPa.

E p a  5N 60 sands with fines

E p a  10N 60 clean NC sands
E p a  15N 60 clean OC sands

[Link]. Correlations for cohesive soils (i.e. Clay)

SPT tests are generally performed in sands and gravels. However, in saturated
clayey soils, due to the proewater pressure development in clay during the
hammering and its influence on the effective stress makes the measured N value
less reliable. As such apart from the hammer energy correction, no other corrections
to N60 are recommended. Approximate correlations with undrained shear strength
(cu) have been postulated - typically cu ≈ 6*N60.

Hara et al. (1974) suggested an approximate correlation below based on SPT tests
carried out on normally to lightly overconsolidated clays (OCR=1-3) from Japan.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.16
Topic 1: Site Investigation and Exploration

𝑐 .
0.24 ∗ 𝑁
𝑝
See also for a very comprehensive summary of all the factors that can affect SPT
tests, and recommendations for interpretation based on the work of others.

4.4.3. Cone Penetration Tests (CPT)

An in situ test, the electric Cone Penetrometer Test (CPT), which can be carried out
in most soils. In this test, a probe is jacked into the ground at a constant penetration
rate. This section briefly describes the test, its advantages and disadvantages, and
its interpretation.

[Link]. CPT probe

The electric CPT, which is an

improvement from the intial mechanical
cone, is widely used throughout the
world to determine the soil profile and
various soil parameters. It basically
involves the hydraulic jacking of a 10
cm² cone with a 60° tip into the soil and
the separate measurement of the cone
and sleeve resistances. The cone
resistance is measured electronically
with strain gauges located behind the
cone. This resistance is usually
identified as qc. In addition a friction
sleeve behind the cone is independently
strain gauged to measure the sleeve
friction fs. The rate of advancement of
cone penetrometers is by default
2cm/sec. Rod lengths are generally
1m, and the continuous penetration
must be halted every 1m to screw on a
new length of rod. As the test is very
rapid, it is an ideal method to get a wide
coverage of site conditions, however, as
there is no sample retrieval, it is
important that it be combined with some drilling unless the site area is well known
from previous investigations nearby.

The type and consistency of the soil is determined primarily by the cone resistance
(the greater the cone resistance, the denser the sand or stiffer the clay), and the
ratio between the friction and cone resistances (called the friction ratio,
(Fr(%)=100*(fs/qc)) In general terms, cohesive soils have a high friction ratio (say
>5%), whereas sands have low friction ratios (<2%). Many researchers have
proposed empirical correlations between cone resistance, friction ratio and soil type.
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.17
Topic 1: Site Investigation and Exploration

Prominent among these are Robertson et al., (1986) and Robertson (2010). Their
results are published widely (e.g. in standard text books and Journals).

In the 1980’s a variety of cone penetrometers have been manufactured which

additionally measure pore pressure using a porous stone located at or behind the
tip of the cone. The purpose of the pore pressure measurements is to allow
improved identification of soil type: in sands, excess pore pressures (u) from
insertion of the cone obviously dissipate rapidly, whereas in more fine-grained
materials, pore pressures will dissipate more slowly. It has been shown that the
location of the porous stone on these piezo-cone (CPTu) devices is extremely
critical, and results for different devices cannot be directly compared.

Original mechanical cone penetrometer systems involved the advancement of the

cone with the sleeve static to measure qc and then the advancement of the sleeve
to rejoin the cone to measure fs. This type of penetrometer is still referred to in texts,
however, such devices are not used any more.

In summary, the CPT has the following advantages:

 It is quick and cheap to use. Many tests to quite reasonable depth may be
accomplished in a day.
 It provides a continuous profile very quickly and therefore is an excellent
method of obtaining detailed information between boreholes or identifying
problem areas.

However it does have the disadvantages as follows:

 No soil sample can be taken.

 Usually should be calibrated alongside boreholes.
 Equipment can be costly.
 Troublesome with gravel and boulders.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.18
Topic 1: Site Investigation and Exploration

[Link]. CPT results

The results of cone penetrometer

INTERPRETED
SOIL PROFILE
testing (qc, fs, Fr(%)) are typically
plotted in real-time on paper tape, and
SLEEVE CONE FRICTION may also be stored digitally. This
DEPTH RELATED TO GROUND LEVEL IN METRES

FRICTION fs RESISTANCE qc RATIO FR

(MPa) (MPa) (%) method therefore allows a very quick
0.3 0.2 0.1 0 2 4 6 8 10 20 1O 5 0 appraisal of subsurface conditions.

-5

-10

-15

-20

-25

-30

LEGEND:
CLAY SAND
SILT GRAVEL

[Link]. CPT correlations

Correlations for soil types

Numerous CPT interpretations have been proposed for soil classifications. Although
newer charts (Robertson, 1990) have been developed based on normalized
parameters (corrected for overburden and water table, often done after the test), the
most popular and simple chart (Robertson, 2010) based on qc and FR(%) which is
easy to use and can be used in real-time during the CPT is given below. In the figure,
ISBT is the soil behaviour index and pa=atmospheric pressure=100 kPa.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.19
Topic 1: Site Investigation and Exploration

qc/pa

Fr (%)

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.20
Topic 1: Site Investigation and Exploration

Correlations for clays

It is common to infer a value of undrained shear strength (su) from the qc values
and the total overburden stress, o. The following expression is used:

where Nk = the cone factor (commonly varies between 10 and 30, and 11 to 19 for
NC clay and may approach 25 for OC clay).

The appropriate value for Nk for a particular site should be determined by correlation
with undrained triaxial tests on undisturbed clay samples.

Correlations for preconsolidation pressure (’c) and overconsolidation ratio (OCR)

are given below.
.
Mayne & Kemper (1988): 𝜎 0.243 ∗ 𝑞 where, qc and ’c are in MPa.

𝑂𝐶𝑅 0.37 ∗ where, o and ’o are total and

effective stresses.

The outcome from above correlations for a specific site must be verified with
appropriate laboratory tests (e.g., Laboratory consolidation test).

Correlations for sands

Kulhawy & Mayne (1990) proposed several correlations for sands as:

For NC sand: 𝐷 % 68 𝑙𝑜𝑔 1

∗

𝐷 % . . ∗ 100 Where, Qc=compressibility factor

(0.91=highly compressible, 1.0=moderately compressible, 1.09=low compressible

sands), pa=atmospheric pressure=100 kPa
𝑞
𝜙 tan 0.1 0.38 ∗ 𝑙𝑜𝑔
𝜎
Many more correlations exist. However, for a given site, a number of these
correlations are to be used and calibrated with laboratory and locally practiced
values for a reasonable confidence.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.21
Topic 1: Site Investigation and Exploration

[Link]. CPT - SPT correlation

A useful correlation between SPT N60 value and qc has been proposed by Kulahwy
& Mayne (1990) as follows:

.
5.44 ∗ 𝐷 Where, D50=Mean particle size (in mm), pa=atmospheric
pressure=100 kPa

4.4.4. Pressuremeter Tests (PMT)

The pressuremeter is a probe which can be inserted in the ground and then inflated.
It was originally developed in 1956 as a relatively sophisticated instrument for
measuring the in-situ strength and elastic modulus of soils and rocks. Although the
equipment is relatively expensive, its use in the field can be very economic with
many tests being made in a working day. There are basically two forms of the
pressuremeter. One is installed in a prebored hole while the other is a self boring
device.

[Link]. Menard Pressuremeter

The Menard
pressuremeter (G-AM
II) probe which is a
cylindrical metal body
with an inner rubber
membrane and outer
protective sheath
mounted so as to
form three
independent cells.
The middle cell is
inflated with water
and the two guard
cells with gas. The
probe is insereted into
a pre-drilled borehole
at a desired depth for
measurements as shown here (Rocktest, 2020). The deformations are measured
by the central measuring cell where conditions of plane strain exist due to the
presence of the guard cells.

The water flow to the measuring cell and gas to the guard cells are regulated via the
control unit fitted with regulators and valves. The volume changes are read directly
from the graduated sight-tube on the reservoir. A coaxial tubing is used to connect
the probe to the control. The inner tube carries the water to the measuring cell while
the space between the two tube walls carries the gas for the two guard cells. This

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.22
Topic 1: Site Investigation and Exploration

method prevents the expansion of the inner tubing which would lead to erroneous
readings of the amount of water injected.

[Link]. Self-boring pressuremeter

The self-boring
pressuremeter is in
principle the same as
the Menard device, but
in addition it has a
specially designed
cutting head at the
bottom end of the
device. This consists of
a rotating cutter which
cuts the soil and this
soil is in turn brought to
the surface by means
of circulating water.

The principle advantage of this device is that the soil disturbance is minimized. For
a Menard pressuremeter it is necessary to drill an oversized hole into which the
pressuremeter is placed; between drilling and installation, the borehole may relax
or even collapse. The self-boring pressuremeter avoids these problems.

[Link]. Pressuremeter operation

For the Menard pressuremeter there are three main testing procedures according
to the soil type allowed.

(a) For free standing soils in which the sides of the borehole do not cave in the
borehole is advanced by the appropriate method and the probe is lowered to
the testing location. Usually tests are conducted at one metre intervals.

(b) For soft clays and granular soils the borehole may be advanced and kept open
with drilling fluid. The test is conducted as above. According to Menard (1957)
the effect of the bentonite drilling mud is negligible, however, others may
reasonably consider this to be an optimistic view.

(c) For the same soft or granular soils a special casing may be used. It consists of
conventional casing which has narrow longitudinal slots machined into it. The
casing will support the borehole walls but also allows the expansion of the
pressuremeter. However it is obvious that the expansion characteristics of the
slotted casing must first be measured, and appropriate correction applied to the
results. The use of drilling mud is preferred.

Self-boring pressuremeters advance to the test depth from the base of a borehole
drilled to just above that depth.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.23
Topic 1: Site Investigation and Exploration

[Link]. Phases of the Pressuremeter test

Once the pressuremeter has been installed at its test location the zero readings are
taken with the pressuremeter sides just touching the soil. The pressure is then
progressively increased and the volume of the cell is measured at each incremental
pressure until the soil fails or the volume of the cavity has expanded by twice of the
original volume (V1). Once done, the proble is deflated and moved to another depth.
The pressure readings are ploteed against the volume volume to obtain effectively
a stress-strain plot, as shown.

There are three basic stages involved in the deformation of the soil as shown in the
figure below.

2Go Stage 1 (up to po) consists of the

recompression of the soil to the in-situ stress
Pressure, p

pl 1
3 before the borehole was made.
2 Stage 2 (po to pf) consists of the elastic
pf
deformation of the soil adjacent to the
measuring cell. This is characterised by a
p straight line plot of the pressure/volume curve.
The elastic modulus E is usually derived from
po 1 V this section of the curve.

Stage 3 (pf to pl) is equivalent to the plastic

Vo V1 V2 V3=2V1 deformation of the soil in which the volume
Total cavity volume, V change increases until at the limit pressure pl
Cavity strain, =dr/ro (%) the soil fails.

[Link]. Determination of elastic modulus from PMT

The pressuremeter may be used to evaluate the properties of most soils and soft
rocks. However, because the test is relatively quick, the analysis is usually based
on undrained condition for cohesive soils and drained condition for granular soils.
Obviously for soils containing a significant portion of silt sized particles or for rocks
containing significant porous fabric such as fissures and joints, it may not be clear
what type of parameter is being derived by the test.

The appropriate elastic modulus (drained or undrained) of the soil or rock is

evaluated from the linear portion of the pressure-volume curve. The modulus is
given by the expression:

𝐸 2 1 𝜐 𝑉 Where, Vm=(V1+V2)/2; =Poisson’s ratio (=0.5 for

undrained case and select appropriate value for drained case, use 0.33 in
desperation)

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.24
Topic 1: Site Investigation and Exploration

The soil modulus depends on several factors (Briaud, 2013) one of which is the
strain level as can be seen in the above PMT pressure-strain plot at the cavity wall.
Based on the theory of elasticity and using the slope of the initial straight line
(dashed red line) of the PMT pressure-strain plot (between stage 1 and 2), the initial
shear modulus (G) can be calculated as

𝐺 Where, strain at the cavity wall==dr/ro(%)=Change in radius/initial radius

of the cavity.

The shear modulus can then be related to Young’s modulus (E) for isotropic material
(often this is not the case for geomaerials) as

𝐸
𝐺
2 1 𝜐
It is important to remember that the elastic properties determined by pressuremeter
tests are lateral values, and that unless the soil is isotropic, the derived values will
be different from the vertical values because of the manner in which soils are laid
down over time. As the elastic values needed for footing settlement calculations are
vertical, these values may be inappropriate. Pressuremeter results are more
appropriate for piles which may be laterally loaded.

[Link]. Determination of strength parameters

The strength of soils may also be estimated from the pressuremeter test result.
However, in simple terms, it is only possible to evaluate cu for undrained cohesive
soils and ’ for drained granular soils. When a soil possesses both c’ and ’ (e.g.
rocks or mixed soils) a variety of complex testing procedures must be undertaken.

For cohesive soils tested under undrained condition, the undrained cohesion (cu)
may be approximated based on the cavity expansion theory (Baguelin et al. 1978)
as:

𝑐 Where, 𝑁 1 𝑙𝑛

The value of Np typically varies in the range 5 to 12, with an average of 8.5. Ohya
et al (1982) correlated Ep with field SPT N60 as follows:
.
For Clay: 𝐸 1930 ∗ 𝑁 Where, Ep in kPa

.
For sand: 𝐸 908 ∗ 𝑁 Where, Ep in kPa

For granular soils, fully drained conditions are assumed and ’ is evaluated.
According to the original Gibson and Anderson (1961) analysis, at failure the ratio
of circumferential stress to radial stress is given by

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.25
Topic 1: Site Investigation and Exploration

 1  sin 
 N [N.B. Not the N as applies to SPT tests]
r 1  sin 

The pressure-volume relationship is given by

/
𝑝 𝐸/2𝑝 1 𝜐

Therefore, by plotting log (V/V) against log p’, a reasonably straight line is obtained
for the plastic region. The slope of this line is ½(1-N) and therefore ′ may be
calculated.

[Link]. Closing comments

While the above simple treatment of the pressuremeter analysis has been
presented, there are still a great number of factors which influence the test results
and which are not fully understood yet.

These factors include the following:

 The influence of soil dilation;

 The influence of tensile stresses in the circumferential direction especially
with regard to tensile cracking;
 The influence of porewater pressures;
 Drainage conditions;
 Soil variability and anisotropy;
 Loading in a horizontal plane.

The pressuremeter was initially seen as the answer to all geotechnical problems,
but over time it has taken a more sensible position as one of a number of less than
perfect tools which are available to the geotechnical engineer. Pressuremeter tests
are used quite extensively for larger projects where the cost of this type of test can
be justified.

4.4.5. Dilatometer Tests (DMT)

The dilatometer is a blade which can be hydraulically jacked into the base of a
borehole, or pushed with a CPT rig. The blade consists of a flat blade 220mm
(length) x 95mm (width) x 14mm (thickness). A thin, expandable 60mm diameter
steel membrane is located flush at the centre of one side of the plate as shown
below. The membrane can be expanded laterally by 1.1mm against the soil. The
pressure A at “lift-off” of the membrane and the pressure B at which the membrane
expands by [Link] in the soil are recorded. In general, at 200mm intervals or
greater this test can be performed. The recored pressures are corrected and

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.26
Topic 1: Site Investigation and Exploration

subsequently combined to give a dilatometer

modulus, ED, a lateral stress index, KD, and a material
index, ID. The material index and dilatomet modulus
can be correlated to determine the nature of the soil
and its unit weight (Schmertmann (1986). Details can
be found in standrad text books.

According to Marchetti (1980), the test can be used to

determine the full range of strength and
compressibility parameters (E, Ko, OCR, cu, ', and
mv). However, it must always be borne in mind that
insertion of the probe causes soil disturbance and
failure, and the derivation of elastic parameters and
peak strength parameters are therefore indirect and
empirically based.

[Link]/flat-
[Link]

4.4.6. Vane Shear Tests (VST)

The vane shear test is a method used predominantly in soft clays to estimate in-situ
shear strength. The cruciform section vane is generally pushed into the soil at the
base of the borehole, however, it can also be pushed from the surface with a
protective sheath. The vanes can be of different widths and lengths to allow a broad
range of soil strengths to be tested. The height of the vane is twice its diameter.
Once in place, the vane is rotated at a constant rotational speed of 6°/minute, and
the maximum torque achieved is recorded.

The applied torque, T, to turn the rectangular vane is related to the undrained
strength of a soft clay, su(VST) by

𝑇 𝑠 ℎ
Where, h is the height of the rectangular vane and d is the diameter. For other types
of vane, such as the tapered vane, a separate form of the above equation is used.
Both the maximum torque and the residual torque (i.e., 5-10 revolutions after then
maximum torque) can be measured in the field and the relevant undrained shear
strength values can be calculated. The caulcuated undrained shear strength value
from the vane is found to be higher and requires to be corrected for actual design
purpose as:

𝑠 𝜆𝑠 Where, λ=Correction factor

Bjerrum (1972): 𝜆 1.7 0.54log 𝑃𝐼%

.
Morris & Williams (1994): 𝜆 1.18𝑒 0.57 𝑃𝐼 5
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.27
Topic 1: Site Investigation and Exploration

. %
𝜆 7.01𝑒 0.57
4.4.7. Field Permeability Tests

Permeability tests include both in-situ permeability tests and permeability tests on
samples retrieved from the boreholes for laboratory testing. Although the in-situ tests
have advantages in that tests encompass the global response of the soil, with soil
fabric relatively undisturbed, interpretation can be difficult when the soil is distinctly
layered, or boundary conditions do not match the simplified model in some other
way.

By contrast, laboratory permeability tests can isolate the permeability of a particular

soil, however, the effect of soil fabric is often lost in the necessarily small samples.

Permeability tests are not an issue with foundations unless the foundations require
an excavation below potential water-table level.

4.4.8. Applicability of Test Methods

The previous sections have described a number of in-situ testing methods that are
available to the geotechnical engineer. This is by no means an exhaustive list of
possible tools, but covers the principle tools used for typical site investigations.
Tabulated below is an extensive range of geotechnical tests and their applicability
to plan for a site investigation work.

Not surprisingly, the self-boring pressuremeter emerges as one of the best

investigation tools in all categories. The piezo-cone devices also rate highly. It must
be remembered that both devices, but particularly the self boring pressuremeter are
highly specialized pieces of equipment that are not readily available, and are
expensive to use. Their use must be consistent with the available budget and the
accuracy of estimation of material properties required in the design project.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.28
Topic 1: Site Investigation and Exploration

4.5. Sampling

Samples fall into two broad categories - disturbed and undisturbed. The key features
of each of these categories will be discussed hereafter.

4.5.1. "Disturbed" Sample

 A by-product of most operations.

 Structure and often moisture content considerably altered from insitu state.
In certain cases, contamination from different strata occur (e.g. flight auger).
 Give some definition of changes in strata.
 Used in laboratory for visual classification and mechanical analysis index
determination (e.g. liquid and plastic limit).

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 1.29
Topic 1: Site Investigation and Exploration

 The amount of disturbance varies. The returned wash from the wash boring
process is extremely disturbed; the sample obtained during an SPT is
significantly disturbed.
 Handling is important. When obtained samples should be placed
immediately in airtight jars, tins, plastic bags etc. labelled to identify the job
number, locality, bore no. depth and date of sampling and sent to lab.

4.5.2. "Undisturbed" Samples

 Samples in which the insitu structure and moisture content have been as far
as possible completely preserved.
 No sample truly "undisturbed" because of stress release on sampling.
 In stiff and over-consolidated clays, sampling not likely to significantly affect
the 'design properties' of soil.
 In soft and sensitive clays, sampling disturbance considerably reduces
measured strength, modulus etc. and increases compressibility.
 In sands, undisturbed samples are rarely attempted;
 The simplest method of obtaining undisturbed samples is pressing a thin
walled tube into the soil i.e. an open drive sampler.
 Sampler should not be driven into the soil but pressed in, preferably at a
constant rate of penetration.
 Open drive samplers commonly around 50 mm and 75 mm inside diameter.
L/D ratios vary from about 6 to 18.
 To avoid disturbing soil the borehole should be cleaned out before sampling
(e.g. with auger or by jetting). Sample ahead of earlier boring. Also maintain
ground water balance. In soft and sensitive soils the thin walled piston
sampler gives best samples. Tube forced into ground with piston blocking
off end. At required depth, the piston is released from ground level and
sampler advanced over the piston. Enables samples to considerable depths
in soft clays without casing. In variable or very hard soils this cannot be
used. In order to obtain "undisturbed" samples, construction of sampler
must be appropriate. Critical parameter is the area ratio which is defined as:

𝐴 % x 100

 Large area ratio gives large disturbance. However, strength requirement

defines minimum area ratio. AS1726 states that area ratio must not exceed
10% with a cutting edge not exceeding 5° to the axis of the sampler.

5. SITE EXPLORATION PLANNING

Depends on many previously mentioned variables such as existing knowledge,

anticipated site conditions and accessibility, type of structure and its sensitivity to
ground movements, proposed means of construction, financial limitations etc.
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.30
Topic 1: Site Investigation and Exploration

In general, it will be necessary to specify at least the following information:

 number and type of holes (bores or penetrometers etc.);

 their required locations and depth;
 frequency and type of sampling in each hole;
 subsequent treatment of samples, handling and transport precautions;
 field and laboratory testing required;
 extent of supervision of drilling and testing.

The table (ICE, 2012) below provides some useful guidance on the planning stage
of a site exploration work.

It should be emphasised however that there should be effective communication

between the field engineer, the design engineer and the client so that should any
modifications to the exploration programme be necessary they can be made without
undue waste of time or confusion.

6. SITE INVESTIGATION REPORTS

It is of considerable importance that the report which is produced at the culmination

of site investigation is complete and comprehensive and conveys the information
required in clear and effective terms.

It is vitally important that all significant aspects of the site investigation are covered
in relation to the proposed development so that the next phase of the development
(probably the design work) may be satisfactorily carried out. Remember that the
report is generally a legal document and any gross misinterpretation of data may
lead to legal proceedings!
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 1.31
Topic 1: Site Investigation and Exploration

The report will also be a record for use in the future should any defect or
unfavourable condition develop during the life of the development. The report may
also be used for assessing any future extensions or modifications to the original
development.

Should any possible source of concern appear during the investigations this should
also be reported and any recommendation as to its solution should be included.

A clear distinction between factual data and interpretations should be maintained

throughout the report.

It is also important that the report contains a clear identification of the site and the
location of the boreholes test pits etc. All borehole logs and tests results should also
be included preferably in standard tabular form. Cross-sections are always a good
way of illustrating the soil stratification.

A good report should include plenty of diagrams and figures which always speak
louder than words.

Remember that the report is the final and probably the only product of the site
investigation and generally since it represents a considerable amount of time energy
and cost it is worth doing well.

7. REFERENCES

Bowles, J.E. (1997). Foundation analysis and design. 5th ed. International ed.
McGraw-Hill, Singapore.

Briaud, J.L. (2010). Ménard Lecture - The pressuremeter test: Expanding its use.
Proceedings of the 18th International Conference on Soil Mechanics and
Geotechnical Engineering, Paris, September 2-6.

Das, B.M. & Sivakugan, N. (2019). Principles of Foundation Engineering. 9th SI Ed.,
Cengage Boston, USA.

ICE (2012). Manual of geotechnical engineering I & II. ICE, London.

Lam, C. & Jefferis, S. A. (2017). Polymer support fluids in Civil Engineering. ICE
publishing, London, UK.

Marchetti, S. (1980). In-situ tests by flat dilatometer. Jnl. Geotechnical Engineering

Div., ASCE, Vol. 106, GT3, March : 299 - 321.

Robertson, P.K. (2010). Soil behaviour type from the CPT: an update. 2nd
International Symposium on Cone Penetration Testing, Huntington Beach, CA,
USA. Volume 2&3: Technical Papers, Session 2: Interpretation, Paper No. 56.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.1
Topic 2A: Introduction to Shallow Footings

Topic 2A: INTRODUCTION TO SHALLOW FOOTINGS

TABLE OF CONTENTS
1. THE MAJOR ISSUES 2
2. FOOTING DESIGN - THE BIG PICTURE 2
2.1 Progressive loading of soils 2
2.2 Bearing Capacity 3
2.3 Settlement 4
2.3.1 Compatibility of settlement and structure 4
2.3.2 Settlement Rate 8
3. FOUNDATION PRESSURE 8
3.1 Gross Pressure 8
3.2 Net Pressure 8
3.2.1 What is Net Pressure? 8
3.2.2 Net Pressure for Bearing - General Principle 9
3.2.3 Net Pressure for Settlement - General Principle 9
3.3 Ultimate Bearing Pressure 10
3.4 Maximum Safe Bearing Pressure 10
3.5 Allowable Bearing Pressure 10
3.6 Design Pressure 11
3.7 A last note - avoiding confusion 11
4. NET PRESSURE-EXAMPLE CALCULATIONS 11
4.1 Short-term excavation in Cohesive Soils- Example 1 11
4.1.1 Net pressure for settlement 12
4.1.2 Net pressure for bearing 13
4.2 Long Term Excavation in Cohesive Soils (both consolidation, and elastic) and Any Term
Excavation in Granular Soils (elastic). 13
4.2.1 Schematic - Long term excavation (cohesive) or short-term (granular) 13
4.2.2 Net pressure for settlement 13
4.2.3 Net pressure for bearing 15
4.2.4 Summary 15
4.3 Surface footing 15
4.3.1 Surface footing 15
4.3.2 Net pressure for settlement 15
4.3.3 Net pressure for bearing 16
4.4 Wide excavations 16
4.4.1 Wide excavations 16
4.4.2 Wide excavation in cohesive soils-short term (consolidation only) 16
4.4.3 Wide excavation in cohesive soils (long term) and any term in granular soils 17
4.5 Impervious excavation under water 17
4.5.1 Impervious excavation 18
4.5.2 Impervious excavation in cohesive soils- Short term (consolidation only) 18
4.5.3 Impervious excavation in cohesive soils (long term), and any term in granular soils 19
4.6 Pervious excavation under water 19
4.6.1 Pervious excavation under water 19
4.6.2 Pervious excavation under water in cohesive soils (short term, consolidation only) 20
4.6.3 Pervious excavation under water in cohesive soils (Long term), and any term in granular
soils 20
4.7 Net Pressure Cases Summary 20
4.7.1 Summary Table 21
4.7.2 General Patterns 21
5. PRACTICAL CONSIDERATIONS 22
6. REFERENCES 22

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.2
Topic 2A: Introduction to Shallow Footings

This introduction gives a brief overview of the major issues that influence the design
of shallow foundations. A major portion deals with various definitions of pressure,
and in particular, the computation of net pressure for bearing capacity and
settlement calculation. That is, it concentrates on the correct calculation of the
disturbing forces. In subsequent sections, the methods for computing the capacity
of the soil, or the compressibility of the soil will be introduced.

1. The Major Issues

“The foundation is that part of a structure which serves exclusively to transmit the
weight of a structure onto the natural ground" (Terzaghi & Peck, 1967).

It follows, therefore, that the foundation is that part of a structure directly in contact
with the soil or rock. i.e., structure = foundation + superstructure. The foundation
must be designed so that ground movements transmitted to the superstructure are
no more that the superstructure can adequately withstand.

This requires:

1. That there is an adequate factor of safety against shear failure of the soil, i.e. the
structure does not break into the soil.

2. That settlement, both total settlement and differential settlement, does not exceed
tolerable limits.

2. Footing Design - the big picture

Soils deform under the application of footing loads. This section will consider the
deformation effects as a foundation element is progressively loaded. This will be
related to two fundamental concepts in footing design: bearing capacity and
settlement.

2.1 Progressive loading of soils

As an example of the typical deformation of a soil under foundation loading, the

behaviour of an isolated footing on a soil and the consequent deformations will be
considered.

0 < P < P1 P1 < P < P2

For 0 < P < P1: the footing will
P = P2
B
B settle in an elastic manner with
B
point A at the base of the
 A footing moving linearly
P2
downwards with increasing
0 P1 PHASE 2 applied load. Point B adjacent
VERTICAL DEFLECTION

PHASE 1 Movement of B
(elastic) (plastic)
to the footing at the ground
Heave

LOAD P surface will experience a

similar but somewhat smaller

deformation pattern.
Settlement

Movement of A

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.3
Topic 2A: Introduction to Shallow Footings

When P = P1: plastic zones will begin to be formed where the stress concentrations
are high i.e., at the edges of the base. These initially small zones will have reached
their maximum shear resistance and will deform plastically, leading to an increase
in the rate of settlement at point A. At the same time, point B will begin to experience
some upward movement or heave.

As P increases above P1 and towards P2, the plastic zones will increase in size and
this will be reflected in the increasing rate of settlement of point A, and the increasing
rate of heave of point B, as the footing experiences more plastic deformation.

Eventually when P = P2, all the soil underlying the footing (or at least all the soil on
the potential failure plane) will have passed into a plastic state so that all shear
resistance is mobilized with no more to develop. The settlement of point A will have
increased dramatically and point B will have experienced considerable heave. The
footing will have exceeded the bearing capacity of the soil. This example illustrates
the two main criteria in foundation design i.e., the bearing capacity and settlement.

2.2 Bearing Capacity

Bearing capacity is
dependent primarily on
soil strength. As can be
seen in the previous
example, the ultimate
strength of the soil is
realised at P2. Obviously
this point must never be
reached in the foundation.

The structure at right has

not been designed with an
adequate factor of safety,
and has suffered a classic
rotational failure typical of
bearing failure. It is
possible that the soil strengths have been underestimated, that the soil is influenced
to a greater depth than has been assumed, or that shaking of the ground by
earthquake temporarily liquefied the bearing sands or silts.

Similarly just before P2 is reached, deformations are still quite large and may be
harmful to the structure. Therefore, the actual design load on the footing must be
well clear of this region of excessive plastic deformation, preferably in the region
below P1. It has been shown that the point P1 is about 1/3 to 1/2 of the ultimate load.
Therefore, in design, a factor of safety is applied to the ultimate load, to give a safe
load. It follows that this value is between 2 and 3 but depends on many variables to
be discussed later.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.4
Topic 2A: Introduction to Shallow Footings

By designing to the above general principles, there is an adequate factor of safety

against structural collapse due to bearing capacity failure.

2.3 Settlement

Settlement is primarily a function of soil compressibility. Settlement is the amount of

movement below the foundation. In the previous example, if the above criterion has
been applied and the foundation has a design load of P1, the settlement under the
foundation will be . Again this settlement must be compatible with the design of the
structure. i.e. can it tolerate the settlement? if not, the design will have to be revised
to meet the limit.

2.3.1 Compatibility of settlement and structure

When considering the amount of settlement acceptable to any given structure, a

number of factors must be considered and these are primarily connected with the
design and components of the superstructure and the purpose for which the
structure is to be put. These factors will be discussed in detail later in the course.
There are basically two forms of settlement: differential and total settlements.
Generally, differential settlement is the most important aspect of settlement.

[Link] Total Settlement

Consider the simple structure

shown to the right. If it were to
experience a large amount of
total settlement only, then the
T
structure would simply be located
at a lower elevation (represented
by dotted lines) than as first
constructed.

Historically, cities have been built around in river delta areas because of obvious
access to shipping transport. Unfortunately (or fortunately for geotechnical
engineers), these river deltas are areas of deep deposits of soft, normally
consolidated and compressible alluvial deposits. Although this may have been no
problem in the old days when structures were mainly simple and light low-rise
structures, for modern high-rise structures, large settlements must be
accommodated or alternative costly solutions (e.g. piling-deep foundation) must be
used. Good examples of cities built in delta areas are Venice and Bangkok.

Anyone visiting the Doge’s Palace in Venice can see the effects that large
settlements have had on this building.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.5
Topic 2A: Introduction to Shallow Footings

Total settlement - Example 1

Another city built away from the sea, but still subject to massive settlements is
Mexico. Settlement of as much as 6 metres are due to the groundwater removal
from the high water content bentonitic clays and volcanic pumices that underlie the
city.

A good example of total

settlement can be seen
in the Mexico Museum
of Fine Arts on the left.
The first floor of the
building has actually
become the basement
due to a gross
settlement in excess of
3 metres since its
construction. Note also
the sloping car park
around the building
which shows the zone of
influence of the
settlement around the Palace.

Total settlement - Example 2

If a structure was only to
undergo uniform total
settlement there would be
little if any effect on the
overall structural stability.
The major effect of total
settlement is where the
structure interfaces with
external structures - e.g.
services entering the
structure. The induced
shear and bending
stresses have in this case
(left) caused rupture of the
pipes.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.6
Topic 2A: Introduction to Shallow Footings

[Link] Differential Settlement

If rather than settling uniformly, a structure experiences a high degree of differential

settlement, one side of the structure would move significantly more than the other
(see the figure below). Such differential movements cause significant structural
distress unless they are specifically accommodated in the structural design.

 = differential
settlement

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.7
Topic 2A: Introduction to Shallow Footings

Differential settlement - Example 1

The best known example of differential

settlement is the Leaning Tower of Pisa (at
right), built in Pisa, Italy between 1174 and
1350. If you look carefully, you can see that
progressive adjustments to the building were
made during construction to correct the lean.
The differential settlement is caused by
variations in ground conditions below the
tower in combination with high applied
foundation stresses and removal of
groundwater from a regional aquifer below
the tower for irrigation. The 18th committee
for the restoration of the tower has
successfully arrested the leaning, and will
(only slightly) reverse it to ensure its safety
without affecting its tourist potential.

Differential settlement - example 2

Differential settlement can also be seen in structural
distress. This is perhaps a rather severe example of
cracking resulting from differential movements in the
foundations of a house. Without close examination it
is not possible to say whether these movements have
been settlements or heaves or a combination of both.
If you keep your eyes open, you will see many less
dramatic examples of such cracking. Look
particularly around door and window frames, where
the corners act as stress concentrators.

Differential settlement can obviously have a serious

effect on the serviceability of the structure or even
affect the overall stability of the structure. Whether
the settlement is acceptable or not would depend on
the structure itself, its design, its function and its
contents. If the structure were glass-sided or
contained sensitive machinery such as printing
presses, the maximum value of  may be as little as 10 mm or even less. If the
structure was only a single storey steel-framed warehouse with flexible design, 
could be satisfactory at greater values, possibly as much as 50-100 mm.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.8
Topic 2A: Introduction to Shallow Footings

Differential settlement - Example 3

As a further example of the variability of allowable settlements, consider an oil
storage tank with a total centre settlement of 1.5 m, and a total edge settlement of
0.75m. Yet this may operate satisfactorily. Obviously settlement must be known so
that the design of piping etc. may be achieved. If, however, the edge settlement
varied between 0.75 m and 1.5 m, then the tank would twist and cause all sorts of
problems including seizure of the floating roof.

Therefore, as we see, settlement tolerances can be extremely variable and will

depend on the structure. These aspects will be covered in more detail in Chapter 3.

2.3.2 Settlement Rate

The rate at which settlement occurs will depend on the soil in question. In granular
(free draining) soils, this settlement will occur immediately on loading of the
foundation. However, with cohesive soils (very slow draining soil), a finite length of
time will elapse between loading and reaching a final settlement (in very fatty clays
it may take up to tens of years to get there).

3. Foundation Pressure

In order to design a foundation against bearing capacity failure or excessive

settlement, it is necessary to determine the net pressure exerted.

3.1 Gross Pressure

The gross pressure exerted by a foundation on a soil is simply the total force
(including the weight of the foundation) acting at the base of the foundation divided
by the area of the foundation base. The gross applied pressure is independent of
the depth of embedment of the foundation for a footing of the same dimensions. In
general, the gross pressure is not used in design of footings.

3.2 Net Pressure

The net pressure exerted by a foundation on a soil can be different from the gross
pressure. Since in soil engineering it is usually the net pressure which is important,
care must be taken in its correct estimation.

3.2.1 What is Net Pressure?

In general terms, the net pressure is that component of the gross pressure which

 for bearing resistance, requires mobilisation of the soil strength to support the
foundation;

 for settlement, contributes to the volume change of the soil due to loading and
hence the movement experienced by the foundation.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.9
Topic 2A: Introduction to Shallow Footings

The net pressure for bearing and settlement are not necessarily the same. The net
pressure for bearing is influenced primarily by foundation geometry. However, the
net pressure for settlement is influenced by four factors which are the geometry, the
type of soil, the time which elapses between initial excavation and final construction,
and the type of settlement (i.e. either consolidation or elastic).

3.2.2 Net Pressure for Bearing - General Principle

Determining the net pressure for bearing is very simple. All that is needed is to
understand the mode in which bearing failure occurs. Both from the schematic in
the previous example and from the various photographs already shown, it can be
appreciated that soil failure involves a rupture that extends below the footing and
emerges, as heaved soil to the side of footing.

The net pressure for bearing therefore involves a difference in the effective pressure
beneath the loaded footing and the effective pressure at the same depth beside the
footing. If the footing is buried a significant distance below the surface, there will be
a surcharge effect next to the footing which will act to reduce the tendency for this
rupture to occur.

Mathematically, this can be expressed as:

𝑞 𝑞′ 𝑞′

𝜎′ 𝛾′𝐷

Where ’ is the effective unit weight of the soil and Df is the depth of soil adjacent to
the base of the footing. You are strongly advised to understand this general principle
and to use the following particular applications only to appreciate how the general
principle is applied.

3.2.3 Net Pressure for Settlement - General Principle

Settlement does not involve a rupture to the side of the footing, but rather a vertical
settlement below the footing. Therefore, the effective pressure beside the footing is
irrelevant to the determination of net pressure for settlement.

Settlement is nothing but a change in the arrangement of soil particles below the
footing into a closer packed arrangement. This closer packing results from the
applied footing load which imposes higher stresses on the soils.

This rearrangement may either be due to the immediate elastic response of the soil
(sand or clay), which is fully reversible on removal of the load, or it may be due to
long-term squeezing out of water in clays below the footing, known as consolidation.

What is important to the calculation of settlement is the difference between the

effective pressure under the footing before the application of the load and the

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.10
Topic 2A: Introduction to Shallow Footings

effective pressure under the footing after application of the load. It is this pressure
difference which will be responsible for settlement.

Mathematically, this can be expressed as:

𝑞 𝑞′ 𝑞′

𝜎′ 𝛾′𝐻

Where ’ is the effective unit weight of the soil and Ho is the depth of soil at the
footing prior to the settlement calculation. As for the net bearing pressure, you are
strongly advised to understand this general principle and to use the following
particular applications only to appreciate how the general principle is applied.

3.3 Ultimate Bearing Pressure

The ultimate bearing pressure (qu) is the value of loading intensity, in terms of either
net or gross pressures (qu net or qu gross), which will cause the soil to fail in shear
loading to give a bearing capacity failure.

3.4 Maximum Safe Bearing Pressure

The maximum safe bearing pressure (qs) is the pressure that a soil will safely
support without the risk of shear failure, irrespective of any settlement that may
result.

In terms of gross stresses: 𝑞

or, in terms of net stresses: 𝑞

Please note that since qu net and qu gross can be different, the respective values of
FoS (Factor of Safety) can differ. In soils the net stress is used, and therefore, the
second of the above two relationships is more usual. However, in any report, it
should be made quite clear which of net or gross stresses is being considered.

3.5 Allowable Bearing Pressure

The allowable bearing pressure (qa) is the maximum allowable load intensity on the
soil. This is not only from a bearing pressure aspect but also from the estimated
amount and rate of settlement, and the ability of the structure to accommodate it.

i.e. 𝒒𝒂 𝒒𝒔
For example, it has been estimated that for a certain structure, qs = 50 kPa.
However, the foundation is to support very delicate and settlement sensitive
machinery. If qs was applied, the estimated settlement may be 50 mm, which is not

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.11
Topic 2A: Introduction to Shallow Footings

acceptable. Therefore, qa would have to be lowered to satisfy the lower value of the
settlement.

It follows that qa is a function of the soil characteristics, the depth and dimensions of
the foundation and the type of structure. It is the maximum design pressure.

3.6 Design Pressure

The design pressure (qd) is the value of foundation pressure used in the final design.

i.e. 𝒒𝒅 𝒒𝒔 𝒖𝒔𝒖𝒂𝒍𝒍𝒚 𝒒𝒅 𝒒𝒔
3.7 A last note - avoiding confusion

The soil engineer is generally concerned with net values while the structural
engineer is often only interested in the gross values. It follows that considerable care
must be taken to make sure that there are no misunderstandings when
communicating design information between the two.

The above definitions do vary from place to place and in practice must be clarified
with those with whom you deal.

4. Net Pressure-Example calculations

The following examples cover a wide range of possible footing configurations, and
are provided only to demonstrate the application of the general principles for
determination of net pressures for bearing and settlement covered elsewhere in this
document. Do not attempt to learn each case separately by rote; the point is to
understand the principles, after which you can quickly compute any case by yourself
especially in the exam or in the design!

4.1 Short-term excavation in Cohesive Soils- Example 1

Consider an excavation of depth H made into a cohesive soil in which the time
between excavation and construction is short. Assume the clay movement is
controlled by consolidation only i.e., the elastic component of clay rebound and
settlement are ignored.

Four stages of the

H
construction sequence will
H q1
be considered as shown in
this figure (left). The unit
A A A
A weight of the soil is .
(a) (b) (c) (d)
Immediately Immediately Foundation Foundation
prior to after excavation placed with fully loaded
excavation and before loading  H loading and settled

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.12
Topic 2A: Introduction to Shallow Footings

4.1.1 Net pressure for settlement

Stage (a): Imagine a small element of soil A, located immediately below the bottom
of a future excavation. This element will have a depth H below the ground level.

Stage (b): Immediately after the excavation is formed, since insufficient time has
elapsed to permit the clay to swell (or experience negative consolidation), element
A will still be at the same depth.

Stage (c): If the foundation is now placed and loaded to H (the pre-existing soil
stress) then the stress at stage (a) are simply being reimposed and the element A
will still be at the same depth. It follows that for these first three stages, the element
will not have moved.

Stage (d): If we now impose the balance of the gross pressure (q1 - H) to fully load
the foundation, the soil beneath the foundation will consolidate, and element A will
move down.

(a) (b) (c)

Elevation

Time The movement of

Depth

E Point A with time is

shown schematically
(d) in the left figure.

From the above, it is clear that the consolidation settlement of the foundation is
controlled by the pressure component (q1 - H), which is the net pressure, and not
by the gross pressure, qgross. In other words,

𝑞 𝑞 𝛾𝐻 𝑞 𝛾𝐻

The net pressure for consolidation settlement could be termed as the difference
between the eventual final effective pressure (f’) and the effective pressure which
existed immediately before foundation placement (o’).

From the above, since insufficient time has elapsed for the porewater pressure to
stabilize and swelling to take place, the initial effective pressure is given by

𝜎′ 𝛾𝐻
The final effective pressure is given by,

𝜎′ 𝑞

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.13
Topic 2A: Introduction to Shallow Footings

From the above: 𝑞 𝜎′ 𝜎 𝑞 𝛾𝐻

The effect of water table is discussed in section 4.5.

4.1.2 Net pressure for bearing

Now the net pressure for bearing resistance will be considered. Whether the soil has
sufficient time to rebound on excavation and then settle on reimposition of the soil
load (Stages (a) to (c)) is unimportant in this case. Any rebound would only change
the depth H by a few millimetres. Since H is in terms of meters, the influence is
negligible. Supposing we have got as far as stage (c) with H applied to the soil.
Since this quantity of pressure is exactly balanced by the soil alongside the
excavation, there is no resultant force trying to displace the foundation. It is only in
stage (d) after the load is increased beyond H towards q1, that strength must be
mobilised. Therefore, the net pressure for bearing is given by,

𝑞 𝑞 𝛾𝐻 𝑞 𝛾𝐻

Which, in this case, is the same as the net pressure for settlement.

Alternatively, the net pressure for bearing could be estimated as the difference
between the eventual final effective pressure immediately beneath the foundation
(f’) and the effective pressure which exists immediately beside the foundation (b’).

From the above: 𝑞 𝜎′ 𝜎 𝑞 𝛾𝐻

4.2 Long Term Excavation in Cohesive Soils (both consolidation,

and elastic) and Any Term Excavation in Granular Soils (elastic).
The net pressures for cohesive soils where the excavation is allowed to stay open
a long time before placement of the footing are the same as for excavations in
granular soils.

4.2.1 Schematic - Long term excavation (cohesive) or short-term (granular)

Foundation
placed Four stages of the
here
construction sequence will be
H
H q1 considered as shown in the
left figure. The unit weight of
A
A A the soil is .
A
(a) (b) (c) (d)
Immediately Immediately Foundation Foundation
prior to before placing placed with fully loaded
excavation foundation  H loading and settled

4.2.2 Net pressure for settlement

Stage (a) represents the at-rest in-situ equilibrium state prior to the excavation.
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2A.14
Topic 2A: Introduction to Shallow Footings

Stage (b): For cohesive soils- there will now be sufficient time after excavation for
the base and hence element A to swell upwards and therefore raise its elevation.
For granular soils, this movement is immediate ie., this swelling would occur
immediately after excavation. Cohesive soils also exhibit some elastic deformation,
and therefore, this would also occur immediately on excavation. Therefore, it is
assumed that time is sufficient to let both elastic and consolidation to fully occur for
all soils.

Stage (c): The foundation is now placed on the base of the excavation at its new
increased elevation and loaded to H. This will cause the element A to return to its
original position.

Stage (d): When the balance of the gross pressure (q1 - H), is applied, further
settlement will occur. With cohesive soils, for the application of both H and (q1 -
H), consolidation settlement will take time whereas elastic settlement for cohesive
and granular soils will be immediate.

The variation of elevation of Point A with time is shown in the figure below.

(a) (b) Since the foundation was

(c)
Time placed at stage (c), H will
Elevation

E cause some settlement

Elastic Consolidation which will be experienced
settlement
settlement
(d) by the foundation. In
addition, (q1 - H) will
cause further settlement
as before. It follows then that the net pressure with regard to settlement for these
conditions comprise all of the gross pressure and H does not come into the
argument.

i.e. 𝑞 𝑞 𝑞

Using the general principle, the effective pressure which existed immediately before
foundation placement (at Stage (c)) was zero since swelling (or elastic rebound) has
occurred. Therefore,

𝜎′ 0
As before, the eventual final effective pressure, f’, is given by,

𝜎′ 𝑞

𝑞 𝜎′ 𝜎 𝑞

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.15
Topic 2A: Introduction to Shallow Footings

4.2.3 Net pressure for bearing

As was argued for the short-term excavation, the fact that rebound and settlement
makes negligible difference to the value of H and therefore, the net pressure for
bearing is the same as in for the short-term excavation case.

𝑞 𝜎′ 𝜎 𝑞 𝛾𝐻

4.2.4 Summary

On the basis of the above discussion it would appear that the net pressure for
consolidation settlement for clays depends on the time lapse between excavation
and construction. If the time is short, then qnet for consolidation is less than qgross by
a value equal to H.

i.e., 𝑞 𝑞 𝛾𝐻

If the time between excavation and construction for clays is long enough to allow
swelling, or for elastic settlements for both clays and granular soils, then qnet is the
same as qgross.

i.e., 𝑞 𝑞

For the net pressures for bearing, it does not matter whether this time is long or
short, but is a function of geometry.

4.3 Surface footing

An isolated footing at the ground surface is a special, although impractical case.

Inevitably, footings will be buried some depth into the founding layer. This section
deals with the net pressures for settlement and bearing for this special case.

4.3.1 Surface footing

q
q2 2
This right figure depicts the geometry of a surface
footing. In this case there is no excavation and
therefore no question of swelling on excavation.

4.3.2 Net pressure for settlement

For net pressures for settlement, since no excavation takes place, o’ and f’ are
quite obviously the same for both short and long term cases. Therefore,

𝜎′ 0 𝑎𝑛𝑑 𝜎′ 𝑞

𝑖. 𝑒. , 𝑞 𝑞

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.16
Topic 2A: Introduction to Shallow Footings

4.3.3 Net pressure for bearing

For the net pressure for bearing, since time does not matter,

𝜎′ 0 𝑎𝑛𝑑 𝜎′ 𝑞

𝑖. 𝑒. , 𝑞 𝑞
4.4 Wide excavations

An isolated footing within a wide excavation is typical of the situation where isolated
column loads bear on the floor of a wide basement excavation. This section deals
with the net pressures for settlement and bearing for this special case.

4.4.1 Wide excavations

This figure below depicts the geometry of a wide excavation.

q3
H

4.4.2 Wide excavation in cohesive soils-short term (consolidation only)

The case of a wide excavation in cohesive soils differs from the narrow excavation
only in terms of net pressure for bearing.

[Link] Net pressure for settlement

Since no swelling (or negative consolidation) takes place, the net pressure for
settlement is given by,

𝜎′ 𝛾𝐻 𝑎𝑛𝑑 𝜎′ 𝑞

𝑖. 𝑒. , 𝑞 𝑞
[Link] Net pressure for bearing

For bearing, as before the net pressure as before is derived from

𝜎′ 0 𝑎𝑛𝑑 𝜎′ 𝑞

𝑖. 𝑒. , 𝑞 𝑞

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.17
Topic 2A: Introduction to Shallow Footings

[Link] Summary

It would appear in this case that even though the excavation is at depth, the term H
does not influence the net pressure for bearing. The reason for this is that the soil
resisting the bearing pressure exists only to the same level as the base of the
foundation. For narrow
Additional
excavations, where the
resistance soil is immediately
adjacent to the footing
right up to the ground
Failure surface, there is an
surface additional resistance (or
With overburden
to resist
No overburden
to resist weight) of H which must
be mobilised.

4.4.3 Wide excavation in cohesive soils (long term) and any term in granular
soils

The cases of long term wide excavations in cohesive soils, or any wide excavation
for granular soils are entirely equivalent with respect to net bearing pressures for
settlement and bearing. In both cases, the net pressure and gross pressure are
equal.

[Link] Net pressure for settlement

For settlement, net pressure is given by

𝜎′ 0 𝑎𝑛𝑑 𝜎′ 𝑞

𝑖. 𝑒. , 𝑞 𝑞
[Link] Net pressure for bearing

For bearing, the net pressure is given by

𝜎′ 0 𝑎𝑛𝑑 𝜎′ 𝑞

𝑖. 𝑒. , 𝑞 𝑞
4.5 Impervious excavation under water

The situation corresponds to an excavation in which water has not had time to enter
the excavation, or is prevented from entering the excavation by some impervious
lining (which would not be common), or most likely, the excavation is being pumped
dry from a sump.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.18
Topic 2A: Introduction to Shallow Footings

4.5.1 Impervious excavation

This right figure depicts a footing q4

Sides impervious,
constructed in an excavation free of water. excavation contains H
no water
The water table is at the ground level.

4.5.2 Impervious excavation in cohesive soils- Short term (consolidation

only)

The important element to remember with these cases is that the excavation does
not contain water, but the base of the footing experiences a buoyant force from the
water table outside the footing excavation.

[Link] Net pressure for settlement

Since no swelling can take place the initial stress is equal to the in-situ overburden
stress (we are now dealing with a buoyant effective pressure),

𝜎′ 𝐻 𝛾 𝛾 𝐻𝛾′
Where, =bulk unit weight of soil, w=unit weight of water and ’=effective unit weight
of soil.

The final gross pressure is less the buoyant uplift (remember Archimedes?)

𝜎′ 𝑞 𝛾 𝐻

𝑞 𝜎′𝑓 𝜎′ 𝑜 𝑞4 𝛾𝑤 𝐻 𝐻 𝛾 𝛾 𝑞4 𝛾𝐻

This is the same result as for short term excavations in clay where no water was
involved. Note that despite this difference, the total overburden pressure (H)
appears in the above relationship and not the effective overburden pressure.

[Link] Net pressure for bearing

For bearing, the net stress is derived from,

𝜎′ 𝑞 𝛾 𝐻 and 𝜎′ 𝐻 𝛾 𝛾

𝑞 𝜎′ 𝜎 𝑞 𝛾 𝐻 𝐻 𝛾 𝛾 𝑞 𝛾𝐻

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.19
Topic 2A: Introduction to Shallow Footings

4.5.3 Impervious excavation in cohesive soils (long term), and any term in
granular soils

As for the short-term cases, the important element to remember here is that the
excavation does not contain water, but the base of the footing experiences a
buoyant force from the water table outside the footing excavation. Whether in
cohesive soils where full rebound may take a substantial time, or for sands where
rebound is elastic and instantaneous, the net pressure will be identical.

[Link] Net pressure for settlement

For settlement, since swelling can take place as well as any elastic rebound, the net
pressure is derived from

𝜎′ 0 𝑎𝑛𝑑 𝜎′ 𝑞 𝛾 𝐻

𝑞 𝜎′𝑓 𝜎′ 𝑜 𝑞4 𝛾𝑤 𝐻

[Link] Net pressure for bearing

For bearing, the net pressure is given by

𝜎′ 𝑞 𝛾 𝐻 and 𝜎′ 𝐻 𝛾 𝛾

𝑞 𝜎′ 𝜎 𝑞 𝛾 𝐻 𝐻 𝛾 𝛾 𝑞 𝛾𝐻

4.6 Pervious excavation under water

The important element to remember with these cases is that the excavation contains
water at the same height as the water table adjacent to the excavation. I can’t think
why one would want to perform construction in this manner except in the case for a
deep pile foundation!

4.6.1 Pervious excavation under water

q5
This figure on the right depicts a footing
Water
placed under water in an [Link] H
water table in the excavation and in the
surrounding ground are at the same level.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.20
Topic 2A: Introduction to Shallow Footings

4.6.2 Pervious excavation under water in cohesive soils (short term,

consolidation only)

As for all other similar cases treated previously, it is assumed that construction of
the footing is too rapid not to allow any significant heave to have occurred. Any
elastic component of the settlement is ignored.

[Link] Net pressure for settlement

For settlement: 𝜎′ 𝐻 𝛾 𝛾 and 𝜎′ 𝑞

Note that there is no buoyant force component this time, if it is assumed that the
foundation is of small volume.

𝑞 𝜎′ 𝜎 𝑞 𝐻 𝛾 𝛾 𝑞 𝛾′𝐻

[Link] Net pressure for bearing

For bearing, 𝜎′ 𝑞 and 𝜎′ 𝐻 𝛾 𝛾

𝑞 𝜎′ 𝜎 𝑞 𝐻 𝛾 𝛾 𝑞 𝛾′𝐻

4.6.3 Pervious excavation under water in cohesive soils (Long term), and
any term in granular soils

Sufficient time is allowed for the soil to fully rebound - a strange requirement for an
excavation in clay under water. I would think substantial softening would occur.

[Link] Net pressure for settlement

For settlement, 𝜎′ 0 and 𝜎′ 𝑞

𝑞 𝜎′𝑓 𝜎′ 𝑜 𝑞5

[Link] Net pressure for bearing

For bearing, 𝜎′ 𝑞 and 𝜎′ 𝐻 𝛾 𝛾

𝑞 𝜎′ 𝜎 𝑞 𝐻 𝛾 𝛾 𝑞 𝛾′𝐻

4.7 Net Pressure Cases Summary

In the above cases, 5 different conditions were examined and the results are
summarised in the net pressure cases summary table below. It is possible to
determine general patterns for net pressures from this table.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.21
Topic 2A: Introduction to Shallow Footings

4.7.1 Summary Table

The following table demonstrates that each particular case should be treated on its
own merits and the respective qnet values should be estimated on the basis of the
geometry of the problem, the soil type involved, the times of construction and the
type of settlement involved.

Net pressures, q net

Conditions For settlement For bearing

(gross pressures include Consolidation - short term Consolidation - long term excavation
self weight of foundation) excavation All elastic settlements

q
1
(i) q1 - H q1 q1 - H

q2
q2 q2 q2
(ii)

q3

(iii) q3 - H q3 q3

q4
q4 - H q4 - w H q4 - H
(iv)

q5
(v) q5 -  'H q5 q5 -  'H

4.7.2 General Patterns

Despite the apparent complexity of the range of results in the net pressure cases
summary table, they can be very much simplified. For example, the condition of
pervious footings under water is most unlikely as foundations are not surrounded by
water, but are generally impervious and extend to the surface in one way or another.

With this simplification, it may be observed that:

(a) For consolidation settlements with short term excavations (which is the usual
situation with clays):
𝑞 𝑞 𝛾𝐻

(b) For elastic settlements for both clays and sands (and the more unusual
consolidation settlements with long term excavations):

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2A.22
Topic 2A: Introduction to Shallow Footings

𝑞 𝑞

(c) For bearing pressure when the excavation comes to the side of the foundation:

𝑞 𝑞 𝛾𝐻

(d) For bearing pressure, when the excavation is significantly wider than the
foundation:

𝑞 𝑞

It is suggested therefore that the results are not memorised but the principles
involved are thoroughly understood.

5. Practical Considerations

It is necessary for engineers to simplify or idealize reality so that workable models

can be developed. The worked examples in this document are examples of such
idealization. Nevertheless, the engineer must also make sure that his/her
idelaizations do not become too far from the reality. Here are a number of practical
considerations which must be taken into account in design:

1. Footings are not weightless - their weight should be taken into account, and it is
unconservative to assume that they simply replace the weight of the removed
soil.

2. Footing excavations are not left open - they are backfilled with either the
excavated material or select backfill. This material adds to the footing pressure.

3. Elastic rebounds of sands or clays occur during excavation and are never
observed. Any rebound is (unconsciously) removed by trimming the excavation
to depth.

4. The elastic settlements due to the footing weight, or any part of the structure
which is cast or erected prior to connection with other structural members will
never be realized. Consolidation settlements that occur prior to connection will
also not be relevant to the structure.

6. References

Terzaghi, K., Peck, R.B. & Mesri, G. (1996). Soil Mechanics in Engineering Practice,
3rd ed. John Wiley & Sons.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.1
Topic 2B: Bearing Capacity of Shallow Footings

Topic 2B: BEARING CAPACITY OF SHALLOW

FOOTINGS

TABLE OF CONTENTS

1. INTRODUCTION 2
1.1 Analytical methods 2
1.2 Historical/Experience 2
1.3 Loading tests 3
2. ANALYTICAL METHODS ( = 0 CONDITION) 3
2.1 Lower Bound Solution 3
2.2 Upper Bound Solution 4
2.3 Prandtl Punch 4
2.3.1 Schematic of Prandtl Smooth Punch 5
2.3.2 Analysis of Prandtl Smooth Punch 5
2.3.3 Rough Punch 6
2.3.4 Rough Punch with  > 0 6
3. APPROXIMATE ANALYSES (C-- CASES OF GENERAL SOILS) 6
3.1 Terzaghi’s Approximate Analysis for Strip Footing 7
3.1.1 Application of Terzaghi Equation to Other Foundation Shapes 9
3.2 Meyerhof's Approximate Analysis 9
3.3 Analyses by Hansen and Vesic 10
3.4 Other Analyses 11
3.5 So which Equation do you use? 11
3.6 Deep Foundations 12
4. RELEVANT SOIL PARAMETERS 12
4.1 Granular Soils 12
4.1.1 What happens when the footing is loaded? 12
4.1.2 Strength parameters for granular soils 13
4.2 Cohesive Soils 14
4.2.1 What do these undrained and effective parameters mean? 14
4.2.2 Do I use undrained or effective? 14
4.3 Other forms of Loading 15
4.3.1 Eccentric Loading 16
5. ULTIMATE BEARING CAPACITY OF LAYERED SOILS 17
5.1 Stronger soil over weaker soil (c’-' soil) 18
5.2 Weaker soil over stronger soil (c’-' soil) 19
5.3 Rigid base at shallow depth 19
6. ULTIMATE LIMIT STATE DESIGN METHOD (ULS) 20
7. REFERENCES 21

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.2
Topic 2B: Bearing Capacity of Shallow Footings
1. Introduction

In this chapter the various

commonly encountered
methods used for the
determination of bearing
capacity for shallow
foundations will be
examined. Essentially this
chapter answers the
question "what is the
maximum pressure which
the soils can withstand for
a given foundation before
the soil will fail?"
There are three principal
approaches which could
be used as follows:

1.1 Analytical methods

There are many different ways of applying analytical methods but all essentially
involve the use of soil mechanics theory to produce a relationship between the ultimate
failure stress the relevant properties of the soil and the geometry of the foundation.
The very complex real situation can only be resolved if a number of simplifying
assumptions are made. It follows that the relevancy of any analytical method is
governed by the extent to which these assumptions are reasonable for the situation
examined. What assumptions are acceptable in one case may not be acceptable to
another case. Therefore, some care must be taken in the selection of the analytical
method used for a particular situation. Since modern engineering practice requires a
reasonably accurate estimation of foundation types and costs long before construction
takes place, the use of analytical methods is by far the most commonly encountered
approach.

1.2 Historical/Experience

Historically the system used for foundation design was based on experience of similar
conditions. The practice for centuries was "if it worked before it should work again. If
it did fall down, make the next foundation wider or deeper". Unfortunately such an
attitude cannot be used quite so literally in today's economic and political climate. The
size of buildings today are very much bigger and more sophisticated and labour is far
from cheap. It follows therefore that foundation failures must be avoided but at the
same time not over-designed. Despite this, experience does play a very important role
in soil engineering. Most soil engineers on the basis of initial site inspections and the
qualitative results of the first few boreholes can make a reasonable estimate of likely
foundation types. However such estimations must ultimately be substantiated by more
detailed analyses.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.3
Topic 2B: Bearing Capacity of Shallow Footings
1.3 Loading tests

This method basically measures the ultimate bearing capacity directly. While such
tests are often performed on difficult soils (e.g. highly variable) they are very costly in
terms of time and effort. This is particularly so with large foundations. Scale models
can be used, but there are a number of scale errors which can be introduced making
the exercise somewhat less attractive. It follows that before such efforts are
undertaken, a reasonable estimation of bearing capacity must be made (so that the
loading system can be designed) and these can be best achieved using analytical
methods. Loading tests are normally used only where overall project capital costs are
relatively high.

2. Analytical Methods ( = 0 condition)

This section on analytical solutions is provided for background and to expand your
knowlede rather than to give you practical solutions to everyday foundation design
problems. Analytical solutions are generally valid for special or particular cases (e.g.
= 0 in this case), and do not provide a means for general solutions. Nevertheless,
some of the more general approaches developed later are based on a combination of
analytical solutions for different special cases.

Analytical solutions can be either upper bound or lower bound. If the upper and lower
bounds of a problem can be determined, the "exact" answer must lie in between.

2.1 Lower Bound Solution

If an equilibrium distribution of stresses can be found which balances the applied load
and nowhere violates the yield criterion the soil mass will not fail or will be just at the
point of failure. Consider an approximate lower
qu = q B
bound solution for a long strip foundation as
1
q q
2
 =0
shown in the figure right.
A A s
ie SOIL IS WEIGHTLESS
In zone 2 the possible principal stresses are
ZONE 1 ZONE 2 plotted.

q q 𝑞 2𝑐
0 0 qA
qA qB

2 Cu 2 Cu
In zone 1 which must be compatible along the
qB 0 boundary between the two zones,
qA qA qA qA STRESSES ON
1 2 SOIL ELEMENT

qB 0
OF EACH ZONE
𝑞 𝑞 𝑞 2𝑐 4𝑐

However although the above is elastically in equilibrium the system is obviously not
kinematically correct. This is therefore a lower bound solution.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.4
Topic 2B: Bearing Capacity of Shallow Footings
2.2 Upper Bound Solution

The soil mass will collapse if there is any compatible pattern of plastic deformation for
which the rate of work of the external loads exceeds the rate of energy dissipation.
qu
This basically means that if a
plausible collapse mechanism is X

chosen and that if the shear strength

of the soil is developed along the DIRECTION R
discontinuities a possible failure may OF MOTION

result. Consider again a long strip Cu

foundation (right figure). Cu

Cu
By taking moments about point X

𝑅
𝑞 𝑅 𝜋𝑅𝑐 𝑅
2
𝑞 2𝜋𝑅𝑐 6.28𝑐
Since we have found the upper and lower bounds the correct answer will be
somewhere in between.

𝒊. 𝒆. , 𝟒𝒄𝒖 𝒒𝒖 𝟔. 𝟐𝟖𝒄𝒖

It is generally this upper bound method which is used, for example slope slability
analysis and a factor of safety applied against total collapse. In this way as long as
there is an adequate factor of safety it does not matter too much about local
overstressing.

It should be noted further that the above methods give no indication of foundation
movements. These however can be calculated by other methods. Also, the upper
bound solution will depend on the correct choice of failure shape. The above gives a
high value of qu but could be improved by choosing a more kinematically admissible
failure surface.

2.3 Prandtl Punch

The Prandtl smooth punch is an analytical solution which was developed for the
theoretical solution of punching into a weightless material. The analysis is developed
for various conditions of the punch base, and material strength as discussed below.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.5
Topic 2B: Bearing Capacity of Shallow Footings
2.3.1 Schematic of Prandtl Smooth Punch

Prandtl (1920) analysed the qu

problem of the penetration of a A E
C
punch into a weightless material.
q qA
The punch was assumed rigid with a I B
frictionless base. Three failure III
zones were considered. B II D

Following on from above  will be

taken as zero.

Zone I- an active failure zone

Zone II- a radial shear zone
Zone III- a passive failure zone identical for  = 0

2.3.2 Analysis of Prandtl Smooth Punch

Considering zone III, on the plane CD

0 45
2𝑐
𝑞 0 𝑐
qA 2
Passing through Zone II:

90 a 𝑞 𝑞 𝜋𝑐
0
qA
(refer to previous example by taking moments
for Zone II)

In Zone I

Minor principle stress = qB

2𝑐
𝑞 𝑞 𝑞 𝑐 𝑞 𝜋𝑐 𝑐 2 𝜋 𝑐 5.14𝑐
2
(Coincidentally the above lies between 4.0cu and 6.28cu)

However, again, this is by considering a weightless soil and having a smooth base on
the punch which is clearly not the real situation.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.6
Topic 2B: Bearing Capacity of Shallow Footings
2.3.3 Rough Punch qu
A C
The previous solution could
be improved by considering 45
D IR E C T IO N
a rough punch. AC in the O F M O T IO N
Cu
right figure now becomes a
slip line. In the same manner
as before it can be shown Cu
that Cu

3𝜋
𝑞 1 𝑐 5.7𝑐𝑢
2
2.3.4 Rough Punch with  > 0

Let us consider the above rough punch case extended to 0, although the soil is still
weightless. By doing this, the solution can be made more general.
qu
The wedge immediately
beneath the smooth footing PASSIVE
ACTIVE
(Zone I) although still an 45 +  /2
45+/2 qA
III 45 -  /2
45-/2
I
active wedge changes shape qB
because of a finite value of . II

Log spiral
The inclination of the sides of
the wedge increase from 45
to (45 + /2). Similarly the angle formed by the passive wedge (Zone III) decreases
from 45 to (45 - /2). It follows that the interconnecting shear zone (Zone II) can no
longer be a radial fan but must become a log spiral which complicates the analysis
considerably.

When the weight is introduced to the soil the mathematics becomes extremely
complex. Indeed for a soil which possesses cohesion (c), an angle of friction () and
weight (), a rigorous mathematical solution has yet to be established. It follows that
empirical solutions must be used to provide reasonable (in terms of complexity and
ease of use) methods of analysis but still providing sufficiently accurate and realistic
solutions.

3. Approximate Analyses (c-- cases of general soils)

Various solutions have been proposed and these will be discussed in this section
below.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.7
Topic 2B: Bearing Capacity of Shallow Footings
3.1 Terzaghi’s Approximate Analysis for Strip Footing

Terzaghi (1943) first extended the rigorous but limited analyses of Prandtl to an
approximate but more general case of a non-surface footing (D > 0) on a c, ,  soil.
B
The frictional nature of qu
the soil is clearly SURCHARGE D
relevant to the strength
 I
contribution of each of 
III 45 - /2
, D and c whereas
none of these latter II
parameters is Log spiral
significantly interactive.
Consequently
Terzaghi's approximation was based on the superposition of three simplified situations
as follows:

For a soil with  and  but c = D = 0:

𝛾𝐵
𝑞 𝑁 .𝑓 𝛾 𝑁
2
For a soil with  and D but c =  = 0:

𝑞 𝑁 .𝑓 𝐷 𝑁 𝑝 Where, po=D and the effective unit weight is to be

considered if soil is saturated.

For a soil with  and c, but  = D = 0:

𝑞 𝑁 .𝑓 𝑐 𝑐𝑁
These situations are specified by the above N parameters with three different failure
surfaces. The superposition is not theoretically valid. However by numerical analyses
Terzaghi was able to provide a combined solution with a little conservatism.

Terzaghi further considered a perfectly rough footing (i.e. allows shear at interface)
and neglected the shear strength contribution from soil above the footing level (i.e. the
soil above the footing level acts only as a surcharge).

Based on Terzaghi's analysis, the bearing capacity for a strip footing on a dry soil is
given by:
𝟏
𝒒𝒖 𝒄𝑵𝒄 𝒑𝒐 𝑵𝒒 𝜸𝑩𝑵𝜸
𝟐
It should be noted that in the figure above,

 Zone I is not a failure zone but stays rigid.

 Zone II is a log spiral shear zone
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2B.8
Topic 2B: Bearing Capacity of Shallow Footings
 Zone III is a passive resistance zone.

This equation is for shallow footings and although originally for continuous strip
footings on the basis of field and experimental evidence it has been extended to cover
rectangular, square and circular footings. Generally, the solution becomes more
conservative with increased depth.

The generalised Terzaghi’s equation for strip footing thus becomes as follows to take
into account the water table should it be present.

𝟏
𝒒𝒖 𝒈𝒓𝒐𝒔𝒔 𝒄𝑵𝒄 𝒑 𝒐 𝑵𝒒 𝜸′𝑩𝑵𝜸
𝟐
𝟏
𝒐𝒓, 𝒒𝒖 𝒏𝒆𝒕 𝒒𝒖 𝒈𝒓𝒐𝒔𝒔 𝒑𝒐 𝒄𝑵𝒄 𝒑 𝒐
𝑵𝒒 𝟏 𝜸′𝑩𝑵𝜸
𝟐
 Nc, Nq, and N are known as the Terzaghi bearing capacity factors and are a
function of  only.
 Strength parameters c and  refer to the soil below founding level (i.e. in the
failure zone)
𝟏
 ’ term in ( 𝜸′𝑩𝑵𝜸 is the effective unit weight of soil below founding depth (i.e.
𝟐
in the failure zone). If the water table is below B depth from the footing base
then ’= otherwise a weighted average ’ is to be used if water table is between
the base and B depth below footing. If the water table is above base then always
use ’.
 ’ term in (𝒑 𝒐 𝜸′𝑫 is the effective weight of soil above the founding depth
(i.e., the surcharge). If the water table is above the founding depth then the
effective stress at founding depth is to be calculated using the effective unit of
the submerged zone and total unit weight of the dry zone. If the water table is
below the base then use ’= for surcharge.
 The N factors (refer to standard text books (e.g., Das & Sivakugan, 2019) for
their values against ) are for long strip footings but may be modified for other
shapes.

The net pressure applied to the underlying soil is the pressure in excess of the
overburden pressure (p’o=’D). As the founding depth (D) for a shallow footing is often
less than 2m the difference between the net and gross ultimate bearing capacities is
expected to be minimal. However, for deep excavations, this may not be the case. In
the above net equation, the unit weight of soil and concrete is assumed to be the same
(though this may not be the reality!).

This solution is strictly applicable to shallow foundations where D/B < 1.0 but in
practice also applied to foundations with D/B > 1.0.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.9
Topic 2B: Bearing Capacity of Shallow Footings
3.1.1 Application of Terzaghi Equation to Other Foundation Shapes

Exact solutions for shape corrections have been obtained. Terzaghi modified his
equations as follows to give suitable approximate solutions for footings other than strip
footings.

Because of the additional resistance to failure at the sides of the footings, the bearing
capacity is increased above that of the strip footing.

[Link] Square

The modified Terzaghi equation for square footings is:

𝒒𝒖 𝒏𝒆𝒕 𝟏. 𝟑𝒄𝑵𝒄 𝒑 𝒐
𝑵𝒒 𝟏 𝟎. 𝟒𝜸′𝑩𝑵𝜸

[Link] Circular

The modified Terzaghi equation for circular footings is:

𝒒𝒖 𝒏𝒆𝒕 𝟏. 𝟑𝒄𝑵𝒄 𝒑 𝒐
𝑵𝒒 𝟏 𝟎. 𝟑𝜸′𝑩𝑵𝜸

3.2 Meyerhof's Approximate Analysis

qu IV
III

 I
45 + /2
II
L og spiral

Meyerhof's equation is of the same general form as that of Terzaghi, but includes so-
called shape, depth and load inclination factors as multipliers to the bearing capacity
factors Nc, Nq and N. Note, however that the numerical values of these factors are
different from those proposed by Terzaghi. Meyerhof proposed factors based on
theoretical solutions for cohesion only, friction only and surcharge only solutions,
which were then combined to give a general solution.

Meyerhof assumed the slightly different failure pattern which is shown above.

 Zone I: Active failure zone.

 Zone II: Log spiral shear zone.
 Zone III: Mixed radial plane shear.
 Zone IV: Free body.

The bearing capacity factors used by Meyerhof are:

Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2B.10
Topic 2B: Bearing Capacity of Shallow Footings
𝜙′
𝑁 𝑒 𝑡𝑎𝑛 45
2

𝑁 𝑁 1 𝑐𝑜𝑡𝜙′

𝑁 𝑁 1 tan 1.4𝜙

The generalized form of the bearing capacity equation proposed by Meyerhof is as

follows:

𝒒𝒖 𝒈𝒓𝒐𝒔𝒔 𝒄𝑵𝒄 𝒔𝒄 𝒅𝒄 𝒊𝒄 𝒑′𝒐 𝑵𝒒 𝒔𝒒 𝒅𝒒 𝒊𝒒 𝟎. 𝟓𝜸𝑵𝜸 𝒔𝜸 𝒅𝜸 𝒊𝜸

Where, the s values are footing shape factors, the d values are footing depth factors
and the i values are load inclination factors. Meyerhof’s recommendations for these
factors can be found in texts such as Das & Sivakugan (2019) and Bowles, J.E. (1996).

Meyerhof's solution differs from Terzaghi's in that the strength of the soil above the
base of the foundation is considered to assist more than just by providing a surcharge.
This gives a more realistic failure pattern, particularly for deeper foundations.

However, while the Meyerhof analysis does generally give far more realistic
predictions of ultimate bearing capacity for D/B > 1 the performance or design of
foundations is governed by both bearing capacity and settlement. Therefore, the
bearing pressures predicted by Meyerhof's analysis, even after the application of a
factor of safety may result in a footing design with excessive settlements.

3.3 Analyses by Hansen and Vesic

Hansen and Vesic were both researchers that developed modified bearing capacity
recommendations based on the original method proposal by Meyerhof. The form of
the generalized bearing capacity equations proposed by Hansen and Vesic are
virtually identical to that of Meyerhof, and their Nc and Nq factors are exactly the same.
However, their N factors and their shape, depth and inclination factors differ. Again,
refer to texts such as Das & Sivakugan (2019) and Bowles, J.E. (1996).

Hansen and Vesic’s generalized bearing capacity equations are:

𝒒𝒖 𝒈𝒓𝒐𝒔𝒔 𝒄𝑵𝒄 𝒔𝒄 𝒅𝒄 𝒊𝒄 𝒈𝒄 𝒃𝒄 𝒑′𝒐 𝑵𝒒 𝒔𝒒 𝒅𝒒 𝒊𝒒 𝒈𝒒 𝒃𝒒 𝟎. 𝟓𝜸𝑵𝜸 𝒔𝜸 𝒅𝜸 𝒊𝜸 𝒈𝜸 𝒃𝜸

In addition to the other factors defined for the Meyerhof equation, the b values are
factors for sloping footing base and the g values are additional factors for sloping
ground.

Their N factors are as follows:

Hansen & Vesic: Nc & N- same as Meyerhof values.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.11
Topic 2B: Bearing Capacity of Shallow Footings

Hansen (1970): 𝑁 1.5 𝑁 1 𝑡𝑎𝑛𝜙

Vesic (1973, 1975): 𝑁 2 𝑁 1 𝑡𝑎𝑛𝜙

3.4 Other Analyses

There are several other analyses in technical papers on bearing capacity. However,
these are often extremely complex and have not been verified sufficiently by
experimental evidence. However until a more adaptable solution is obtained, the
Terzaghi solution will probably be continued for some time as it is applicable to virtually
all soils based on standard accepted testing techniques.

The following facts should be borne in mind when considering the development of
newer solutions.

 The engineering profession is notoriously conservative and even if a more

accurate method of bearing capacity estimation were developed it would take
many years or decades to make an impact on everyday professional practice.

 Following on from above, engineers tend to use methods which are simple to
apply provided the results are acceptable. Any improved method is likely to be
more complex and therefore less easily applied. Also confidence in its
application would be lacking and therefore there would a strong tendency to
use existing methods which are well proven even though they may be
conservative.

 Bearing pressure estimations do not appear to be a major problem as is

evidenced by the very few cases of major bearing capacity failures. Therefore
the methods currently used (although conservative) do work. Settlement
however has caused innumerable problems and more effort should be made in
this direction rather than in the bearing capacity direction.

 The accuracy of any expression or analysis relating ultimate bearing capacity

to soil properties is governed not so much by the analysis itself and the
assumptions made but much more by the accuracy and the reliability of the soil
properties used. Soils are immensely variable and unpredictable and the
problems of deriving relevant and representative soil strengths are immense.
Therefore, if any effort were to be made at improving predictions of bearing
capacity these would be far better directed at the determination of soil
properties rather than the analysis itself.

3.5 So which Equation do you use?

That is to assume that there is a correct answer, which there is not. As in all
engineering, our mathematical models are just convenient ways of (hopefully)
estimating the actual behaviour of complex systems. Because of the indeterminate
nature of soils, the actual behaviour is probably more difficult to determine than for
most other (structural) materials.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.12
Topic 2B: Bearing Capacity of Shallow Footings
I don’t particularly mind which method you use (even those not listed here), as long as
you can justify the use of the method by reference to a textbook or paper. The use and
comparison of more than one method would be favourably looked upon!

For what it’s worth, Bowles (1996) has the following opinion:

Bearing capcity eqn. Best for

Terzaghi Cohesive soils where D/B1 or for a quick

estimate of qult to compare with other methods.
Do not use for footings with moments and/or
horizontal forces or for tilted bases and/or sloping
ground.
Hansen, Meyerhof, Vesic Any situation which applies depending on user
preference or familiarity with a particular method.

Hansen, Vesic When base is tilted; when footing is on a slope or

D/B1

3.6 Deep Foundations

For D/B > about 5, the foundation is considered as deep and the equations discussed
above are inadequate. This problem will be discussed later in the course when we
examine deep foundations.

4. Relevant Soil Parameters

In order to determine the bearing capacity of a soil, it is obviously important to choose

appropriate strength parameters. Here are some guidelines for cohesive and granular
soils.

4.1 Granular Soils

Granular soils behave fundamentally differently from cohesive soils because of their
more open and free-draining structure. This section will deal with the type of strength
parameters which are appropriate, and the way in which these parameters are
obtained.

4.1.1 What happens when the footing is loaded?

In order to understand what strength parameters to use for a sand, consider what
happens when the footing is loaded. As the footing is loaded, either progressively
during construction or in service as live or wind loads need to be resisted, the stresses
below the footing will increase. This stress increase will tend to compress the sand
structure, moving particles closer together and reducing the voids between particles.

This compression of the structure happens instantaneously. If there is any water in the
voids or pores between the sand particles, the structure is open enough that this can
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2B.13
Topic 2B: Bearing Capacity of Shallow Footings
be squeezed out immediately to accommodate the new denser packing of the grains.
So, no load is transferred to the water and pore pressures therefore no change of pore
pressure (u. All load is carried by the sand particles, and the closer packing means
increased inter-particle stresses (i.e., increased effective stress though =’ here).

Because of the increased inter-particle stresses, the sand becomes harder to shear.
This is because in order to shear, the particles have to ride up and over each other,
which they are being prevented from doing. The shear strength therefore increases
with the applied footing stresses.

This increase of strength, , with increase of normal stress, , is reflected in the
frictional part of the shear strength equation below.

Δ𝜏 Δ𝜎𝑡𝑎𝑛𝜙′

Because of their free draining nature there is normally no excess porewater pressures
present under a foundation placed on granular soils at any time. Therefore, it is the
drained or effective strength properties which are relevant to the calculation of bearing
capacity on sands and gravels. These properties consist of a finite value of ' with c' =
0. (It may be of interest to note than for cuts in granular soils the same effective
strength properties also apply because excess porewater pressures are not involved).

The main exception to the above general rule is when there are major dynamic or
shock loadings (e.g. earthquakes) applied to a granular soil. Load under these
conditions is applied so rapidly that there is insufficient time for drainage to occur and
liquefaction may occur as effective stresses are thrown onto the porewater pressure.
For such a situation it may be appropriate to deal with the problem in terms of
undrained strength parameters. However this is not a problem that will be considered
here.

4.1.2 Strength parameters for granular soils

In order to obtain strength parameters for sands to use in bearing capacity (or any
other) analysis, it is necessary to perform in-situ tests. Can you guess why?

Contrary to the situation which exists with cohesive soils it is very difficult to obtain the
angle of internal friction of granular soils directly for use in bearing capacity analyses.
As will be discussed at length later it is virtually impossible to obtain an undisturbed
sample for laboratory testing. A disturbed sample is of little use as its properties will
not be the same as the soil as it was in its in-situ undisturbed state. Therefore, in-situ
tests have to be relied on. Unfortunately in the great majority of these tests with the
exception of perhaps the pressuremeter test the angle of internal friction cannot be
measured directly but must be inferred from correlations with the measurement made
by the test technique.

Discussions about the Standard Penetration Test (SPT) and Cone Penetration Test
(CPT) can be found in the notes on Site Investigation, as well as numerous other texts.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.14
Topic 2B: Bearing Capacity of Shallow Footings
4.2 Cohesive Soils

4.2.1 What do these undrained and effective parameters mean?

Essentially sands and clays behave in a very similar manner. They both respond to
applied load by compressing, and adopting a more closely-packed and therefore
stronger structure. The main difference is that because of their open structure, sands
can do this immediately, whereas clays can take a long time to adopt the new more
compact structure.

Let’s assume that the clays have any amount of time to rearrange and adopt the
smaller structure. Remember the term consolidation? Just as for sands, the strength
of the clays is (virtually) dependent only on the applied stresses i.e., it is predominantly
frictional in behaviour, and is proportional to the effective stress. Why is the effective
stress? Because it is the total stress less the soils own buoyancy which determines
the level of stress between particles. So for clays in the long term,

Δ𝜏 Δ𝜎𝑡𝑎𝑛𝜙′

Note: just as for sands, but mind you, quite often there is a small cohesive component
as well - the so-called “true cohesion”, c’.

What does happen in the meantime as this clay is consolidating? Initially as the load
is applied, if the clay is below the water table, there is no way that the clay particles
can get closer together because there is all the pore water in between. As the clay
particles can’t get closer (i.e., no change in effective stress), there can’t be any
increase in strength. But the extra load applied must be taken by something and in this
case, it is the water which takes the extra load, and the increased load in the water
causes the water pressure (or pore pressure) to increase.

So, here we have a situation where no matter how much load we apply to the clay, in
the short-term there can be no increase in clay strength - that is the strength is stress
independent. What is the stress-independent part of the strength equation? - see if
you were right.

Now, assuming that the footing is loaded, and the pore pressure has increased, what
is going to happen? There will be a pore pressure difference below the footing with the
surrounding pore water, which will tend to drive water away from the footing area. As
this water is driven out (over time), the result will be consolidation of the clay and a
strength increase that will depend on the stress applied and the time allowed for
dissipation of pressures. However, in the long-term, as described above, the strength
of the clay will be simply stress-dependent, or frictional.

4.2.2 Do I use undrained or effective?

In general terms when construction takes place on cohesive soils the undrained
parameters (i.e. the undrained cohesion cu and the undrained angle of internal friction
u which is conventionally taken as zero) are usually used. The reason for this is that
the time taken for construction can be very roughly of the same order of for the

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.15
Topic 2B: Bearing Capacity of Shallow Footings
dissipation of excess porewater pressures in clays. This means that when a structure
is built there may be excess porewater pressures existing under the foundations which
will only dissipate some time after construction has been completed. It follows that the
strength of the soil will not have changed a great deal from its strength prior to
construction. Therefore so that safety is assured it is the undrained parameters (i.e.
the parameters which existed before drainage takes place) that are considered the
most critical and are therefore used in design.

When a structure is placed on a clay consolidation settlement can go for several years
and while this settlement will cause a strength gain in the clay it may not be realised
until long after the structure has been put into service. While it is the larger drained or
effective strength parameters which will eventually apply once drainage has been
completed they would not be the more critical. Nevertheless, there are situations
where some allowance can be made for strength gain during construction. However
this would require a great deal of data on rates of porewater pressure dissipation,
consolidation, settlement and strength characteristics which for real soils is difficult to
determine.

Therefore in summary for structures placed on cohesive soils it is the undrained

strength properties (cu and u = 0) which generally apply.

It might be interesting to compare this situation with the construction of cuts in clay
materials where because stresses are removed (rather than applied as is the case
with structures) the most critical time is some time after construction when the induced
negative porewater pressures have dissipated to produce a clay of lower strength. It
follows that for this situation the most critical time is a long time after construction.
Therefore, it is the effective (or drained) strength properties (c', ') which should be
used to check ultimate stability rather than the undrained properties.

Unlike granular soils, cohesive soils are relatively easy to sample and test in the
laboratory. Thin-wall tubes (or u-tubes) are used for taking undisturbed samples of
clays or soils with some clay content. Tubes vary from about 50mm to 80mm diameter,
and wall thickness is about 1mm. The tubes are sealed at either end to retain moisture
content during storage.

For cohesive soils because of the many methods available both in the laboratory and
in the field for the determination of properties, it is comparatively easy to obtain
reasonably accurate undrained strength properties for bearing capacity estimations.

4.3 Other forms of Loading

Unfortunately, columns do not always sit nicely in the middle of footings (perhaps
because of architectural constraints), and columns or walls may also impose lateral
loads or moments to the foundations from environmental loading such as earth
pressure or wind loading. Here are some guidelines for accounting for these problems.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.16
Topic 2B: Bearing Capacity of Shallow Footings
4.3.1 Eccentric Loading

It is relatively common to find foundations loaded such that the line of action of a
vertical load does not coincide with the centroid of the foundation area. This is
particularly common at edges of buildings near the site boundary or when axial load
and moment work together. A method for analyzing this situation is given in the
following.

[Link] Stress distribution under eccentric loading

When loading is eccentric, a rigid foundation will no longer give rise to a uniform
pressure distribution on the supporting soil. The variation in contact pressures are
given in the figure to the right.

For 𝑒 c
e
P
𝑃 6𝑒 𝑃 6𝑒
𝑞 1 𝑎𝑛𝑑 𝑞 1
𝐵𝐿 𝐵 𝐵𝐿 𝐵

For 𝑒
q min
4𝑃
𝑞 𝑎𝑛𝑑 𝑞 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒 q max e<
B
3𝐿 𝐵 2𝑒 6

Note: For 𝑒 , the value for qmin becomes

zero and with 𝑒 , the value for qmin B
q max e>
becomes negative meaning tensile stress 6

development. As soil cannot resist tension,

separation between foundation and soil
would take place. In practice, it is expected to limit eccentricity to 𝑒 .

[Link] The Meyerhof effective area method

The method for handling this situation in design is referred to as Meyerhof's (1953,
1963) effective area method. It
e1
consists of adjusting the area of P
the foundation so that the vertical
load becomes a central load. This
is illustrated for one-way L
eccentricity in the diagram.
P
For an eccentricity of e1 in the L B
direction the new length becomes

𝐿 𝐿 2𝑒 2e 1 L/2 L/2

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.17
Topic 2B: Bearing Capacity of Shallow Footings
The contact stress q for rectangular distribution is considered to be

𝑃 𝑃
𝑞
𝐴 𝐵𝐿′ 𝐵 𝐿 2𝑒

For two-way eccentricity (with e2 in the B direction, not shown here) the new area
becomes

𝐴 𝐵𝐿 𝐿 2𝑒 𝐵 2𝑒

Therefore, with two-way eccentricity, the uniform contact stress (q) becomes

𝑃 𝑃
𝑞
𝐴′ 𝐿 2𝑒 𝐵 2𝑒
This method ensures that the design load employed is always developing stress less
than qmax so that the safety is assured.

The above effective footing dimensions can be used to determine the ultimate net
bearing capacity (q’u) of an eccentrically loaded footing based on Meyerhof (1963)
method discussed earlier. The important issue to be considered here is the effective
L’ and B’ when calculating the shape factors (s). However, when calculating the depth
factors (d) the original width (B) is to be used. Finally, the ultimate load can be
calculated as Pu=q’uA’.

5. Ultimate Bearing Capacity of Layered Soils

The theories put forward so far deal with generally simple and somewhat idealised
problems. In reality soils are not as simple as this and soil properties often change
dramatically with depth. Therefore, depending on the particular problem, modifications
have to be made for its solution. Some very general applications will be studied here
although often a literature search may have to be undertaken should soil conditions
merit it. There are many specific solutions to be found in the literature.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.18
Topic 2B: Bearing Capacity of Shallow Footings
5.1 Stronger soil over weaker soil (c’-' soil)

The bearing capacity of a

thick bed of stronger soil
(qt) is much higher than
the underlying weaker
soil (qb). At ultimate load, Ca
Stronger soil Df
a punching shear failure
will take place within the ( 1, c’1, ’1 )
top stronger soil followed
by a general failure in the Stronger soil (1, c’1, ’1)
bottom weaker soil. This Weaker soil
is shown in the right (2, c’2, ’2) qt>>qb
figure (modified from (a) Thin layer (b) Thick layer
Meyerhof, 1974).
However, if the stronger soil is much thicker then the ultimate bearing capacity will be
soley dependent on this layer.

The ultimate bearing capacity for a strip footing placed in the above layered system
(qu) can be approximated as (Meyerhof & Hanna, 1978):

2 𝐶 𝑃 𝑠𝑖𝑛𝛿
𝑞 𝑞 𝛾 𝐻
𝐵
Where, Pp=passive resistance on the punching surface, =inclination of the passive
force with horizontal, Ca=c’aH=adhesive force, c’a=interface adhesion which depends
on c1.

The above equantion can be simplified to the following form:

2𝑐′ 𝐻 2𝐷 𝐾 𝑡𝑎𝑛𝜙′
𝑞 𝑞 𝛾 𝐻 1 𝛾 𝐻 𝑞
𝐵 𝐻 𝐵
Where, Ks=punchinh shear coefficient=(q2/q1, ’1), q1 & q2=bearing capacities of a
surface strip footing placed on thick beds of stronger and weaker soils separately,
qt=the ultimate bearing capacity of a strip footing placed in top stronger
soil=cNc+p’oNq+0.5BN.

The variations of Ks and ca with q2/q1 ratios can be found in Das & Sivakugan (2019)
or Meyerhof & Hanna (1978).

The above equation for strip footing can be extended to a rectangular footing by
incorporating the effect of L and B on the punching volume as:

𝐵 2𝑐′ 𝐻 𝐵 2𝐷 𝐾 𝑡𝑎𝑛𝜙′
𝑞 𝑞 1 𝛾 𝐻 1 1 𝛾 𝐻 𝑞
𝐿 𝐵 𝐿 𝐻 𝐵

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.19
Topic 2B: Bearing Capacity of Shallow Footings
You should also apply shape factors when calculating the ultimate bearing capacities
for the top (qt) soil as Df the founding depth and bottom (qb) soil as (Df+H) as the
founding depth. The depth factors are not considered here.

5.2 Weaker soil over stronger soil (c’-' soil)

For a footing is placed on a thin weaker soil underlain by a stronger soil (i.e., q2/q1>>1)
then the top weaker soil can fail by squeezing out. The ultimate bearing capacity in
this case can be estimated using a semi-imperical relationship as proposed by
Meyerhof (1974) below.

𝐻
𝑞 𝑞 𝑞 𝑞 1 𝑞
𝐻

Where, H=thickness of the top weaker layer and Hf=depth of failure surface beneath
footing in a thick bed of top weaker layer (For vertical loading and strip footing, Hf≈B
for loose sand and clay, Hf≈2B for dense sand; for circular footing, Hf≈B for all soils)
(Meyerhof, 1955). The depth of failure surface decreases significantly with increased
load inlcinations.

qt and qb are to be estimated using a thick layer of each soil separately and using the
Meyherhof bearing capacity equation with shape factors ONLY. The founding depth
(Df) remains the same.

5.3 Rigid base at shallow depth

For a strip footing, if a rigid rough base is located at a depth below the footing, which
is within the depth of the failure zone then the full development of the failure surface
will be restricted. Mandel & Salencon (1972) developed an ultimate capacity equation
for strip footing in such case as:

𝑞 𝑐𝑁 ∗ 𝑝′ 𝑁 ∗ 0.5𝛾𝐵𝑁 ∗
Where, N*=modified bearing capacity factors which are dependent on friction angles
and H/B ratios (refer to Das & Sivakugan, 2019). For H>B, the effect of the presence
of a rigid base for lower friction angle (<30o) could be ignored and the general bearing
capacity factors (N) could be used.

For the rough rectangular and circular footing with a rough rigid base, the above strip
footing equation was modified to incorporate the shape factors as:

Rectangular/circular footing on sand (c’=0): 𝑞 𝑝′ 𝑁 ∗ 𝑠 ∗ 0.5𝛾𝐵𝑁 ∗ 𝑠 ∗

Where, 𝑠 ∗ 1 𝑚 ; 𝑠∗ 1 𝑚 are modified shape factors, m1 and m2 vary

with H/B ratios and friction angles (refer to Das & Sivakugan, 2019).

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.20
Topic 2B: Bearing Capacity of Shallow Footings
√
Square footing on saturated clay ('=0): 𝑞 𝜋 2 𝑐 𝑝′ (for
H<0.707B)

6. Ultimate limit state design method (ULS)

The previous sections deal with the ultimate bearing capacity of a shallow footing
placed on thick or layered soil deposit. However, in real design, one has to calculate
the size of the footing based on the design load and soil properties so that adequate
factor of safety (FoS) against the shear failure is ensured. In practice, a combined FoS
between 2 to 3 is used on the soil ultimate capacity to determine its allowable capacity,
where a lower value of FoS is associated with higher confidence in soil properties. The
allowable footing load carrying capacity is then compared with the actual load applied
on the footing (ie., Fallowable ≥ Fapplied) to ensure safety against soil bearing failure.
However, this does not address the issue with tolerable settlement which has to be
estimated to ensure that the settlement does not exceed the limiting value during the
serviceability period. This will be discussed later.

Recent design codes use the ultimate limit state (ULS) design approach where the
uncertainities associated with estimating structural loads and soil properties can be
handled separately based on risk classifications. This allows more control on the
design side, for example, if more investigation is done to obtain soil properties then
geotechnical risk can be reduced or FoS can be lowered.

The ULS design equation as you have been practicing in many civil engineering units
can be written as:

𝐸 𝜙 𝑅 .

Where, Ed = Design action effects (actual load multiplied by appropriate load factors);
Rd,ug = design ultimate geotechnical strength;g = geotechnical reduction factor which
depends on how much a designer is confident with the design soil properties. Typically
g can vary between 0.35 and 0.65 (low value for high risk design or low confidence in
soil properties or less investigation has been carried out). Field and lab investigation
may improve the confidence level with soil properties thus less reduction of the
geotechnical capacity. You should follow AS5100.3-2004 Tables 10.3.3(A) & (B) for
further details.

The basic ultimate limit states used in checking strength (Ed) are given below while
many other load combinations can be found in the standards (AS1170, Section 4.2.2).
The long-term and combinations factors can be found in Table 4.1 (AS1170).

Ed = 1.35G [permanent action only]

= 1.2G + 1.5Q [permanent and imposed action]
= 1.2G+1.5lQ [permanent and long-term imposed action]
= 1.2G + Wu + cQ [permanent, wind and imposed action]
= 0.9G+Wu [permanent and wind action reversal]

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2B.21
Topic 2B: Bearing Capacity of Shallow Footings
= G + Eu + EQ [permanent, earthquake and imposed action]
= 1.2G+Su+cQ [permanent acition, actions given in Clause 4.2.3 &
imposed action]
Where, G=permanent action; Q=imposed action; Wu= ultimate wind action;
Eu=ultimate earthquake action, E=earthquake combination factor, C=Combination
factor, l=Long-term factor (often the values equal to C for distributed imposed action,
check Table 4.1-AS1170).

7. References

Bowles, J.E. (1996). Foundation Analysis and Design. 5th edition. McGraw-Hill.

Das, B.M. & Sivakugan, N. (2019). Principles of Foundation Engineering. 9th edition.
Cengage, Boston, USA.

AS1170 (2002). Structural design acitions. Part 0: General principles. Australia/New

Zealnd Standards. Standards Australia.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.1
Topic 2C: Settlement of Shallow Footings

Topic 2C: SETTLEMENT OF SHALLOW FOOTINGS

TABLE OF CONTENTS
1. INTRODUCTION 3
2. STRESSES IN SOIL MASS 3
2.1 Stress from a vertical point load 4
2.2 Stress below a uniformly loaded circular area 4
2.3 Stress below a line load 6
2.4 Stress below a uniformly loaded flexible strip area 6
2.5 Stress below a horizontal strip load 7
2.6 Stress below a uniformly loaded rectangular area 7
2.7 Stress below a uniformly loaded irregular area 9
3. SETTLEMENT MECHANISMS AND COMPUTATIONS 10
3.1 Immediate settlement 10
3.1.1 Essential features of immediate settlement 10
3.1.2 Appropriate parameters for immediate settlement estimation 10
3.1.3 Immediate settlement based on the theory of elasticity 11
3.1.4 Immediate settlement due to a single concentrated load 11
3.1.5 Immediate settlement due to a uniformly loaded circular area 12
3.1.6 Immediate settlement due to a uniformly loaded rectangular area over an infinite layer 13
3.1.7 Immediate settlement due to a uniformly loaded rectangular area over a finite layer 14
3.1.8 Immediate settlement of shallow foundation in layered soil 15
3.1.9 Typical modulus values for preliminary settlement calculations 16
3.1.10 Typical values of Poisson’s ratio for estimating immediate settlement 17
3.2 Consolidation settlement 17
3.2.1 What is primary consolidation? 17
3.2.2 Calculating consolidation settlement - Oedometer method 18
3.2.3 Typical consolidation parameters 19
3.2.4 Skempton and Bjerrum's correction for overconsolidated soils 20
3.2.5 Other methods for computing consolidation settlement 23
3.3 Secondary consolidation 24
3.3.1 What is secondary consolidation? 25
3.3.2 Calculation of secondary consolidation 25

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.2
Topic 2C: Settlement of Shallow Footings

3.3.3 Typical secondary compression index values 26

4. CORRECTIONS TO TOTAL SETTLEMENT 26
4.1 Rigidity Correction 26
4.2 Depth Correction 27
5. TIME RATE OF SETTLEMENT 27
5.1 Principle of Estimation 28
5.1.1 Excess pore water pressure distribution 28
5.1.2 Time constant (T) 29
5.1.3 The coefficient of consolidation 29
5.1.4 The maximum drainage path 29
5.1.5 Time-settlement curves 30
5.2 Problems in Practice 30
5.2.1 Influence of radial drainage 30
5.2.2 Influence of fabric 31
6. SWELLING AND GROUNDWATER CHANGES 32
6.1 Swelling of Excavations 32
6.2 Ground Water Table Changes 32
7. OVERALL SUMMARY 33
8. LIMITING VALUES OF TOTAL SETTLEMENT 33
9. SETTLEMENT BASED ON FIELD TESTS 33
9.1 Settlement based on SPT 34
9.2 Settlement based on CPT 34
9.3 Settlement based on Pressuremeter test 36
9.4 Settlement based on Plate Bearing Tests 36
10. EFFECT OF VIBRATIONS 37
11. DIFFERENTIAL SETTLEMENTS 37
11.1 Settlement sensitive structures 38
11.2 Settlement tolerant structures 38
11.3 Reasons for differential settlement 38
11.3.1 Discernible differences in boundary conditions 38
11.3.2 Differential settlement inherent in total settlement 39
11.4 Methods of Reducing Differential Settlement 40
12. OTHER CAUSES OF GROUND MOVEMENT 40
13. REFERENCES 41

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.3
Topic 2C: Settlement of Shallow Footings

1. Introduction
Apart from bearing pressure, the other major design consideration for shallow
footings is settlement. Excessive settlement (primarily differential settlement) can
cause a number of problems. Obviously structural distress can be caused
sometimes catastrophically with additional loads being transferred onto other loaded
foundations possibly enough to cause bearing failure.

Settlement is caused by application of increased stress levels to the soil. Before

imposing a load on a soil, the soil has come to some sort of equilibrium state, where
the soil particles can quite happily bear the weight of the overlying soil.

On applying an increased load to the soil through a footing, the soil is subjected to
increased stress levels, which cause the particles to rearrange in a denser structure
that can accommodate these higher stresses. The volume of voids between the soil
particles decreases, and this decrease is reflected in settlement at the soil surface.
Of course, this rearrangement is only caused by changes in the stresses that the
soil skeleton feels - the effective stresses, rather than the change in total stress,
which includes pore water pressure. So, to compute the settlement of a soil, the
effective stresses in the soil before and after application of the new load must be
determined.

For sands, the rearrangement of soil particles is virtually instantaneous, whether the
sand is above or below the water table. If it is below, the sand structure is so open
that water is not impeded from being squeezed out as the sand grains move to their
denser structure. Sands also respond elastically to applied stress (as long as it is
not excessive). If load is taken off, the sand structure will simply revert to its previous
looser state.

For clays, the settlement is not instantaneous, because water travels slowly through
a clay structure - voids are not continuous, and there are electrochemical attractions
etc. There is a small component of elastic rearrangement, however, the bulk of the
volume change occurs over weeks, months or years as the water is displaced from
the more highly stressed areas. This rearrangement of particles is commonly called
consolidation.

Incidentally, with both sands and clays, this rearrangement of particles in a denser
structure is accompanied by an increase of soil strength.

2. Stresses in soil mass

As pointed out before, in order to determine the amount of settlement, it is necessary
to compute the changes of stress in soils. This section describes various theoretical
solutions for elastic stress distributions which will later be used to calculate the
elastic settlement.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.4
Topic 2C: Settlement of Shallow Footings

2.1 Stress from a vertical point load

The stresses due to a vertical point load are in themselves quite irrelevant - point
loads are not applied to soils.
Rather, loads are applied through
footings which distribute the load Q
to an allowable level. However,
this solution can be integrated over
finite areas to determine stresses 
due to real footings. Boussinesq R Z
(1885) derived the expressions for
normal and shear stress at any
point considered inside a perfectly  z
elastic semi-infinite, weightless r
medium. Since when estimating
the settlement it is primarily the
change of vertical stress which is  r 
of importance, the equation for z
is presented here.

3𝑄 1
∆𝜎 /
2𝜋𝑧 𝑟
1
𝑧
The above equation cab be simplified to

∆𝜎 𝐼 Where, 𝐼 /

Iis known as the stress influence coefficient, the values with (r/z) can be found in
standard text books.

2.2 Stress below a uniformly loaded circular area

In order to find the stress directly below the centre of a uniformly loaded circular
area, Boussinesq's equation for z must be integrated over the area.

Consider the small surface element below loaded with q/unit area. The load on this
area can be considered as a point load and is given by q r d dr.

The stress increment (d) at the centre of the circular area at a depth z can be
determined using Boussinesq equation as:
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.5
Topic 2C: Settlement of Shallow Footings

3 𝑞𝑟𝑑𝜃𝑑𝑟 1 q a
𝑑𝜎 / d
2𝜋𝑧 𝑟
1
𝑧
r
dr
For the entire loaded area, the increase of
vertical stress (z) at the centre at depth z z
can be found by integrating the above
equation.
 z
⎡ ⎤
1
∆𝜎 𝑞 ⎢1 /
⎥ 𝑞𝐼
⎢ 𝑎 ⎥
⎣ 1 ⎦
𝑧
Where I is the stress influence factor for circular loaded area and the values can
be found in standard textbooks (Das & Sivakugan, 2019) in terms of (z/a, r/a).

Stress Influence Design Chart – Circular loaded area

The stress influence values (I) z

I = q
below the centre of a uniformly 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
loaded flexible circular area are 0
shown in the figure on the right. 1

If the change of vertical stresses 2

at other points not under the
3
centre of the area are required,
the process of integration would 4
be very difficult. In these cases, Z /a
other methods are used. These 5
will be discussed later. 6

10
Influe nce fa ctors for the stre ss be low the ce ntre
of a unifor m ly loa de d fle xible circula r a re a .

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.6
Topic 2C: Settlement of Shallow Footings

2.3 Stress below a line load

The change of stress (z) below a vertical q=load/unit length

line load of infinite length on a semi-infinite
soil mass can be calculated as: x

2𝑞𝑧
∆𝜎 z
𝜋 𝑥 𝑧
A (x,z)
z

2.4 Stress below a uniformly loaded flexible strip area

The equation for line load can be further extended to determine the change of
vertical stress at a point below the flexible strip load where the load has a finite width
(B) and infinite length (ie., the case of a wall footing).

By considering a small q=load/unit area

strip of area and treating
it as a line load, the x
increase of stress can be B/2
computed at a point. This B/2
incremental stress can
be integrated over the
limits –B/2 to +B/2 to z
obtain the change of
stress (z) for the finite
width loaded strip as: x A
z
For point A (x≥B/2):

𝐵
𝑞 𝑧 𝑧 𝐵𝑧 𝑥 𝑧
4
∆𝜎 𝑡𝑎𝑛 𝑡𝑎𝑛 𝑞𝐼
𝜋 𝐵 𝐵 𝐵
𝑥 𝑥 𝑥 𝑧 𝐵 𝑧
2 2 4

Where, arctangents are in radians.

For point A (0≤x≥B/2), the magnitude of 𝑡𝑎𝑛 becomes negative and a term 
is to be added to get the solution.

The values of stress influence factor (I) for a uniformly loaded flexible strip footing
in terms of (2z/B, 2x/B) can be found in Das & Sivakugan (2019).
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.7
Topic 2C: Settlement of Shallow Footings

2.5 Stress below a horizontal strip load

The change of vertical stress at a B

point A (x.z) for a horizontal flexible
strip load on a semi-infinite soil mass x
q/unit area
can calculated as (Harr, 1966).

z

2𝐵𝑞𝑥𝑧 A (x,z)
∆𝜎 z
𝐵
𝜋 𝑥 𝑧 𝐵 𝑧
2

2.6 Stress below a uniformly loaded rectangular area

Based on the work of Newmark, it is possible to determine the change of vertical

stress under the corner of a uniformly loaded flexible rectangular area (B=width,
L=length).
U n ifo rm s tre s s , q
Below the corner considered the
change of vertical stress (z) is
given by B L

∆𝜎 𝑞𝐼 z


z
Where,

1 2𝑚𝑛 𝑚 𝑛 1 𝑚 𝑛 2 2𝑚𝑛√𝑚 𝑛 1
𝐼 . 𝑡𝑎𝑛
4𝜋 𝑚 𝑛 𝑚 𝑛 1 𝑚 𝑛 1 𝑚 𝑛 1 𝑚 𝑛

𝑩 𝑳
Where, 𝒎 𝒂𝒏𝒅 𝒏
𝒛 𝒛

In the above influence factor equation, the arctangent must be a positive angle in
radians. If it becomes negative (m2+n2+1<m2n2) then a term  should be added to
that angle.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.8
Topic 2C: Settlement of Shallow Footings

These stress influence values at the corner of a rectangular loaded area are
presented graphically below. The chart presents I as a function of the parameters
m and n. Please note m and n are interchangeable in this chart.

0.26
I m = 3 .0 m = oc
A rea covered m = 2 .5
0.24 m = 2.0
w ith uniform m = 1.8
norm al load, m = 1.6
0.22 q m = 1.4
m = 1 .2
nz
x m = 1 .0
0.20 y
mz
z m = 0.9

 z = qI m = 0 .8
0.18 q.I m = 0.7
z
0.16 m = 0.6

N ote: m and n are interchangeable m = 0.5

0.14

0.12 m = 0 .4

0.10
m = 0 .3

0.08
m = 0 .2
0.06

0.04 m = 0 .1

0.02
m = 0.0
0
0.01 2 3 45 0.1 2 3 4 5 1.0 2 3 45 10
V E R T IC AL S T R E S S BE LO W A C O R N E R
O F A U N IF O R M LY LO AD E D F LE X IBLE
R E C TAN G U LAR AR E A.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.9
Topic 2C: Settlement of Shallow Footings

2.7 Stress below a uniformly loaded irregular area

When dealing with a foundation of irregular shape, the stress equations become too
difficult for use for conventional design purposes. Newmark however developed a
method to evaluate the change of stress. This method effectively uses a graphical
integration.
NEWMARK CHART FOR VERTICAL STRESS
The method
involves a Newmark Influence Value 0.005
chart which is
placed over a scale
drawing of the
loaded area. The
scale of the area is
such that the
reference line on the
chart ab represents
the depth z to the
same scale.

The centre of the

chart is placed on
the point below
which the stress is
required and the
number of squares
including parts
enclosed in the
loaded area
counted. The
change of vertical a b
stress is then given
by

∆𝜎 𝑞∗𝑁∗𝐼

Where, q= applied stress; N=the number of squares or parts thereof and I=the
influence value given on the chart.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.10
Topic 2C: Settlement of Shallow Footings

3. Settlement mechanisms and computations

There are three types of settlements: immediate (elastic) settlement, primary
consolidation settlement and secondary consolidation settlement. These are
described in the following sections.

3.1 Immediate settlement

This section describes the essential features of immediate settlement, and the
methods of computation.

3.1.1 Essential features of immediate settlement

Immediate settlements occur immediately on loading a foundation for all types of

geotechnical materials.

For relatively impervious cohesive soils the immediate settlement takes place at
constant volume if the soil is saturated (because the vertical settlement is
accompanied by a compensatory horizontal expansion for  = 0.5). As there is no
volume change (V=0), this settlement is elastic and fully recoverable. For cohesive
soils, the volume change takes place over an extended period of time and is known
as consolidation settlement.

For granular soils, which are generally fast draining, the same processes take place
i.e. undrained elastic settlement with no volume change and consolidation
settlement. However, with granular soils, the consolidation settlement occurs so
rapidly that it is not possible to distinguish between the elastic and consolidation
components. Therefore, for granular soils, all settlement is lumped into the one
category of immediate settlement.

3.1.2 Appropriate parameters for immediate settlement estimation

Immediate settlement for all soil types are estimated on the basis of elasticity
provided the elastic properties of the soil E and  can be evaluated from the site
investigation. Usually, this is possible with cohesive soils, which can be sampled
and tested in a laboratory. For granular soils, however, sampling, in particular
undisturbed sampling is much more difficult, if not impossible in most cases. In-situ
testing techniques do not usually provide elastic parameters but require the use of
empirical methods for settlement estimation as will be discussed later. Some other
methods, such as the pressuremeter do yield elastic parameters. Therefore, the
elastic methods discussed in this section can only be used if we have prior
knowledge of the elastic modulus and to a lesser extent the Poisson's ratio. This
latter value can usually be assumed with sufficient accuracy.

As discussed above all settlement with granular soils is immediate by virtue of the
fast drainage of such soils. The parameters used should reflect this characteristic.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.11
Topic 2C: Settlement of Shallow Footings

Therefore, the drained modulus (E') and the drained Poisson's ratio (') are
appropriate for granular soils.

For cohesive soils, which are relatively slow draining, the immediate settlement
should be estimated using the undrained modulus (Eu), and the undrained Poisson's
ratio (u). This latter value is 0.5 if the soil is fully saturated.

N.B. Since this type of settlement is considered to occur instantaneously it follows

that should any overburden be removed rebound will occur instantaneously.
Therefore, gross construction stresses should be used.

3.1.3 Immediate settlement based on the theory of elasticity

In order to calculate immediate settlement from surface loads on an elastic medium,

it is necessary to determine the vertical strain in the soil. In the previous sections,
the computation of stresses due to different surface loads were established. These
expressions can be manipulated to compute the vertical strain.

From the theory of elasticity, the vertical strain at a point within a loaded medium is
given by

Δ𝜎 𝜐
Δ𝜀 Δ𝜎 Δ𝜎
𝐸 𝐸
Where, z = change of vertical stress; x and y = changes of stresses in the
two horizontal directions.

3.1.4 Immediate settlement due to a single concentrated load

By placing the expressions for z,

x=r, and y= into the strain Q
equation, the vertical strain at a given
point may be calculated.

The expression reduces to: R Z
𝜀 𝑐𝑜𝑠 𝜓 3𝑐𝑜𝑠 𝜓 2𝜐
 z
r
For the vertical displacement at the
surface, the expression for the settlement

is given by:  r

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.12
Topic 2C: Settlement of Shallow Footings

𝜌 𝜀 𝑑𝑧

By changing variables to r and : 𝑡𝑎𝑛𝜓

∴𝑧 𝑟𝑐𝑜𝑡𝜓

Differentiating the above, we get, 𝑑𝑧

𝑟𝑑𝜓 𝑄1 𝜐
∴𝜌 𝜀
𝑠𝑖𝑛 𝜓 𝜋𝑟 𝐸
3.1.5 Immediate settlement due to a uniformly loaded circular area

This is an extension of the result of the

previous case. a d
For the incremental area shown, the
settlement at the centre can be calculated as
follows. r
dr
Load on incremental area dq = q r d dr

Settlement due to incremental area at centre,

𝑞 𝑟 𝑑𝜃 𝑑𝑟 1 𝜐
∴ 𝑑𝜌  z
𝜋𝑟 𝐸
Settlement due to whole of loaded area can be obtained by

2 a q
   (1  2 ) d dr
0 0 E

2q (1  2 )a
 
E
It can be similarly shown that the settlement under the edge of a circular load area
is given by

4q (1  2 )a

E
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.13
Topic 2C: Settlement of Shallow Footings

Special cases of the above occur in perfectly elastic soils with  = 0.5.

15
. qa 3qa
Centre settlement  and Edge settlement 
E E
Also, for any point at a depth zo below the centre of the circular area, the settlement
from beneath that point is given by

2𝑞𝑎 1 𝜐
𝜌 /
𝐸 𝑧
1
𝑎
Using the above expression, the settlement of a layer of finite thickness may be
calculated.

3.1.6 Immediate settlement due to a uniformly loaded rectangular area over

an infinite layer

This tends to be an extremely complex mathematical exercise. However, by

considering the problem in a manner similar to the stress distribution under a
rectangular areas, we can use an influence factor I.

The settlement under the corner of a flexible rectangle under uniform q was
proposed by Schleicher (1926).

𝑞𝐵
𝜌 1 𝜐 𝐼
𝐸

Where, 𝐼 𝑚 𝑙𝑛 𝑙𝑛 𝑚 𝑚 1 and m1=L/B

The principle of superposition should be used for locations other than at the corners.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.14
Topic 2C: Settlement of Shallow Footings

3.1.7 Immediate settlement due to a uniformly loaded rectangular area over

a finite layer

The most common method for B

computing settlement of a q
finite layer due to a flexible A
surface rectangular loaded
area is due to Steinbrenner Compressible
(1934). The settlement at the E layer H
surface of a finite layer could considered
be approximated by the B
difference between the
surface settlement of an infinite layer and the settlement of the infinite layer at the
depth of the actual finite layer base. The settlement at the corner of a foundation for
a layer of thickness H, is expressed in the same generalized form as for an infinite
layer, however, the influence factor, Icorner is expressed as a function of two
variables, F1 and F2.

𝑞𝐵
𝜌 1 𝜐 𝐼
𝐸

Where, Ι 𝐹 𝐹

Values of F1 and F2 were calculated in graphical form as functions of L/B and H/B
as given below in addition to mathematical expressions.

1 1 𝑚 1 𝑚 𝑛 𝑚 𝑚 1 1 𝑛
𝐹 𝑚 𝑙𝑛 𝑙𝑛
𝜋 𝑚 1 𝑚 𝑛 1 𝑚 𝑚 𝑛 1

𝑛 𝑚
𝐹 𝑡𝑎𝑛
2𝜋 𝑛 𝑚 𝑛 1

m1=L/B and n1=H/B

For saturated clays (n=0.5): Icorner = F1

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.15
Topic 2C: Settlement of Shallow Footings

Values of F1 ( ) and F2 ( )
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

L/B = 1
2
Depth factor d = H/B

L/B = 2 F1
4
L/B = 5 F2

6
L/B = 5
L/B = 10
8
L/B = oo L/B = 2 L/B = 10
L/B = 1 L/B = oo
10

Influence values for settlement beneath the corner of a uniformly loaded

rectangle on an elastic layer (Depth D) overlying a rigid base

3.1.8 Immediate settlement of shallow foundation in layered soil

The elastic settlement below the corner of a shallow footing on a layered soil can be
estimated by considering the theory of superposition. The layer immediately below
the base of the footing is assumed to have a rigid base and the settlement at the
corner of a uniformly loaded rectangular area with a finite thickness (H1) and
modulus (E1) can be calculated. For the second layer, it is assumed that layer 2
exists from the base to a depth (H1+H2) with modulus E2 and having a rigid base.
This assumption allows us to calculate the elastic settlement at the corner of a
rectangular loaded area. However, in this calculation, one has to assume the (H1,
E2) soil as an extra layer which in reality does not exist. Thus to estimate the net
effect of (H2, E2) we have to subtract the settlement effect of (H1, E2) from the
(H1+H2, E2). This concept can be extended to subsequent layers if needed. This
method of superposition is shown in figure below for a two layer case.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.16
Topic 2C: Settlement of Shallow Footings

3.1.9 Typical modulus values for preliminary settlement calculations

The modulus of soil widely varies with tests, strain level, age, compositions, soil
disturbance, soil consistency, stress history etc. It is hard to estimate field modulus
from routine laboratory tests. There is no single modulus value for soils rather a
range is more appropriate. As such in settlement calculations, if the modulus is
approximated from correlations then a range of settlement values is to be estimated.
For approximate calculation, modulus values of a select group of soils are tabulated
below and the values need to be cautiously used as the range is often too large
(Bowles, 1996). It is always desirable that one should use a number of possible
values estimated from a range of tests, such as the values obtained from the
laboratory, field and practical experiences.

The following empirical correlations can be used to estimate modulus of soil (Es)
from undrained shear strength (su) of clay.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.17
Topic 2C: Settlement of Shallow Footings

For normally consolidated sensitive clay: Es=(200 to 500)su

For normally consolidated insensitive and lightly over-consolidated clay: Es=(750 to

1200)su

For heavily over-consolidated clay: Es=(1500 to 2000)su

Soil Es, MPa Soil Es, MPa

Clay Sand
Very soft 2-15 Silty 5-20
Soft 5-25 Loose 10-25
Medium 15-50 Dense 50-81
Hard 50-100 Sand and gravel
Sandy 25-250 Loose 50-150
Silt 2-20 Dense 100-200

3.1.10 Typical values of Poisson’s ratio for estimating immediate settlement

Typical values of Poisson’s ratio for common soil types are tabulated on the right
(Bowles, 1996). It is to be noted that for saturated undrained clay, the value of
Poisson’s ratio (u)=0.5 i.e., no change of volume.

Soil type Poisson’s ratio ()

Most clay soils 0.4-0.5
Saturated clay soils 0.45-0.5
Cohesionless soils-medium and dense 0.3-0.4
Cohesionless soils-loose to medium 0.2-0.35

3.2 Consolidation settlement

The following sections describe the phenomenon known as primary consolidation,

or just consolidation, and describe methods used to estimate this quantity.
Consolidations settlements can be estimated with a reasonable degree of accuracy,
as long as care is taken with sample retrieval and testing, and three-dimensional
effects are taken into account. By contrast, the rate of primary consolidation is often
poorly estimated, and generally occurs more rapidly in practice than in the
laboratory.

3.2.1 What is primary consolidation?

Consolidation settlement occurs in all soils, but can only be isolated in fine grained
cohesive soils where the necessary volume changes occur over an extended period.
It takes place after immediate settlement and may take weeks or even years to
complete. The volume changes result in plastic deformations to occur, and this type

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.18
Topic 2C: Settlement of Shallow Footings

of settlement is therefore only partially recoverable. Typically, consolidation

settlements are the most significant component of settlement for clays.

In non-cohesive (granular) soils, this form of settlement does not take place as a
time dependent mechanism because of its generally free draining nature. However,
it does take place immediately and therefore it is combined with the immediate
settlement.

Consolidation settlement is fundamentally different from immediate settlement for

cohesive soils in that a volume change in the soil does take place with soil particles
being forced closer together and pore water (or pore air) being expelled.

3.2.2 Calculating consolidation settlement - Oedometer method

This method makes use of the results of the

z
conventional Oedometer test, where the
consolidation parameters of the soil are
measured. It then becomes necessary to 
compute the stress changes within the soil
mass. The stress changes are computed
using a Boussinesq-type approach f - Final effective
assuming elasticity. stresses

The important parameter for consolidation z

settlement calculation is the net effective
stress change in the soil. Usually, the i - Initial effective
settlement is calculated for the soil by stresses
dividing into a number of sub-layers and the
final total settlement found from the sum of individual sub-layer settlements. There
is no absolute rule for choice of sub-layer thicknesses, however, the thinner the sub-
layer the more accurate will be the answer, as the Boussinesq variation of stress
change with depth is non-linear. Of course, using thinner sub-layers will carry the
disadvantage of more computations. You should extend the calculation to a depth
of 4 times the foundation width, or to a layer at which sub-layer settlements are no
longer significant. Sub-layers thicknesses should be thinner close to the footing
base (say, 0.5m or 0.25B, whichever is the smallest) and can increase with depth.

Two alternative parameters are often derived from the results of an Oedometer test.

One is the coefficient of volume compressibility (mv):

Δ𝑒
𝑚
Δ𝜎′ 1 𝑒
Where, e=change of void ratio; ’=increase of effective stress; eav=average void
ratio during consolidation.
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.19
Topic 2C: Settlement of Shallow Footings

Using this parameter, settlement of a sub-layer of thickness, H, is computed as:

𝜌 𝑚 Δ𝜎′Δ𝐻

The total consolidation settlement can be obtained by summing up all sublayers:

𝜌 𝑚 Δ𝜎 Δ𝐻

The second is a pair of indices known as the compression index (Cc) and
recompression index (Cr):

𝐶𝐻 𝜎′ 𝐶𝐻 𝜎′ Δ𝜎′
𝜌 𝑙𝑜𝑔 𝑙𝑜𝑔
1 𝑒 𝜎′ 1 𝑒 𝜎′

The recompression index (Cr) is used (1st part of the eqn. only) for final stress levels
≤ preconsolidation pressure ('c). The compression index is used (2nd part of the
eqn. only) for initial effective stress levels (’o) ≥ preconsolidation pressure ('c).
Where the initial stress is less than the preconsolidation pressure, and the final
stress exceeds it, both indices must be used in conjunction to determine the total
consolidation settlement.

3.2.3 Typical consolidation parameters

As mv is a stress-dependent parameter, it is not possible to give typical values for

this parameter. Laboratory Oedometer test is to be performed over the change of
stress range expected in soils and represented values of mv are to be estimated
from the e-log(p) plot.

Bowles (1996) summarizes many correlations for Cc for different soil types (Table
2-5) based on parameters such as the soil liquid limit (wL). Actual laboratory data
are always preferred, however, in the absence of such data, typical values may be
inferred from these correlations.

Bowles (1996) suggests that the recompression index (Cr) is typically 10~20% of
Cc i.e.

𝐶 ≅ 0.05 𝑡𝑜 0.1 𝐶
Thus the settlement in an overconsolidated soil will be significantly less than those
generated in a normally consolidated soil.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.20
Topic 2C: Settlement of Shallow Footings

3.2.4 Skempton and Bjerrum's correction for overconsolidated soils

It is often found that with normally consolidated clays the conventional Oedometer
based methods of consolidation settlement estimation provided a reasonable result.
However, with overconsolidated clays, the conventional method seemed to
overestimate consolidation settlement, particularly when the consolidating layer was
thick relative to the width of the applied load.

Skempton and Bjerrum (1957) proposed a more accurate method by considering

the geometry and stress history for overconsolidated clays. The method is still far
from rigorous but endeavours to take a more realistic view of consolidation
settlements with overconsolidated clays.

In the conventional method the amount of consolidation settlement for any given
layer is provided by the relationship

𝜌 𝑚 Δ𝜎 Δ𝐻

Where, H is the thickness of the sub-layer and  is the change in vertical effective
stress.

However, for consolidation settlement, it is the change in induced porewater

pressures (u), which causes consolidation and not strictly speaking the change in
vertical effective stress (). Therefore, the consolidation settlement of a layer
should be given by

𝜌 𝑚 Δ𝑢Δ𝐻

As the Oedometer test subjects the soil sample to one-dimensional consolidation,

the Oedometer settlement can correctly be given by

𝜌 𝑚 Δ𝜎Δ𝐻

Where, the change in total applied stress () is the same as the change of effective
stress (). The two equations can only become the same when  = u. This holds
only when the Oedometer test and the consolidating layer in-situ have the same
boundary conditions.

Since the soil sample in an Oedometer is confined in a ring of steel, the sample
cannot experience lateral yield - all void ratio change must result in a vertical
movement. It follows that for the test to be a model of full scale behaviour, the in-
situ soil must similarly not experience lateral yield. This is approximated only when
the loaded area is very wide compared with the thickness of the consolidating layer.
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.21
Topic 2C: Settlement of Shallow Footings

When this is not the case, lateral yield occurs in-situ. Some of the void ratio change
is accommodated in these lateral movements, and the Oedometer test will
overestimate surface settlements.

The change in porewater pressure due to the change in principal total stresses can
be described by the following:

Δ𝑢 𝐵 Δ𝜎 𝐴 Δ𝜎 Δ𝜎
When no lateral yield occurs in-situ: Δ𝜎 Δ𝜎
and for a saturated soil with B = 1: Δ𝑢 Δ𝜎
Therefore, the Oedometer test is a reasonable model of in-situ conditions.
However, when lateral yield occurs: Δ𝜎 Δ𝜎

Therefore, for overconsolidated clays, where A can be significantly less than 1 (see
table below), from the above pore pressure equation

Δ𝑢 Δ𝜎

The Oedometer method will overestimate settlement because it is based of 1

rather than the smaller value of u.

It follows that for an overconsolidated clay which undergoes lateral yield:

𝜌 𝑚 Δ𝑢Δ𝐻 𝜌 𝑚 Δ𝜎 Δ𝐻
It would appear that the difference between the prediction of consolidation
settlement on the basis of the Oedometer test and the actual consolidation
settlement for overconsolidated clays is a function of the geometry of the problem
and the relevant value of A.

Type of clay A (for elastic stress range)

Very soft sensitive clay 1
Normally consolidated clays 0.5-1
Overconsolidated clays 0.25-0.5
Heavily overconsolidated sandy clays < 0.25

On the basis of the above,

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.22
Topic 2C: Settlement of Shallow Footings

oed   mv H 1
but c   mv H u
z
or c   mv u dz
0
  3 
since u  1A  (1  A)
 1 
z   3 
then c   mv 1 A  1  Adz
0  1 
Relating c to the conventional value of oed

c    oed
z   3 
0 mv . 1A  (1  A)dz
 1 
where  z
o mv 1dz

In order to simplify the above it has been assumed that mv and A are constant.

Then   A  (1  A)
z
  .dz
for   0z 3
0 [Link]

Therefore,  only depends on the geometry of the problem. Skempton and Bjerrum
calculated values for  for various values of z/B for circular and square footings.

z/B 0 1/4 1/2 1 2 4 10 

_________________________________________________________________
circular footing 1.0 0.67 0.50 0.38 0.40 0.28 0.26 0.25

square footing 1.0 0.74 0.53 0.37 0.26 0.20 0.14 0
_________________________________________________________________________

Where, z is the depth to the compressible layer.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.23
Topic 2C: Settlement of Shallow Footings

It follows from the above that settlement calculated from the traditional method for
an overconsolidated clay is overestimated although for a normally consolidated clay
(A = 1), the values agree fairly well.

Generally, although shape affects the  factors, typical values of  are as follows.

Clay Type Approx 

_________________________________________________________________
Very sensitive soft clays around 1 may be greater
Normally consolidated clays slightly less than 1
Overconsolidated clays about 0.5
Heavily overconsolidated sandy clays about 0.25, possibly higher

For further discussion, see Skempton and Bjerrum (1957) and Parry (1968).

3.2.5 Other methods for computing consolidation settlement

The one-dimensional Oedometer method, with or without correction for

overconsolidated clays, as given above, is the most common approach to
computation of consolidation settlements. However, it must be emphasised that
consolidation is generally a three dimensional problem (four, if we include time) and
it follows that a one-dimensional approach can never be considered accurate.
Unfortunately, when working with three dimensions, the analysis of settlement
becomes extremely complex and it is not used to any great extent in commercial
practice. However, as these methods become more refined, they may become more
common and practically useful.

The two methods mentioned below both have in common, very detailed laboratory
testing procedures and a fairly sophisticated analysis. It is important to realize that
the benefits offered by more sophisticated approaches can only be realized if soil
sampling and testing are of the highest standard. Any sample disturbance will
negate the benefits of these methods.

The methods are described for information only.

[Link] Lambe's stress-path method

This method involves the stressing of a representative laboratory sample along the
same stress path (both vertical and lateral) as the soil would experience in the field.
First, the in-situ stress state must be determined (not trivial) and replicated. The
changes in the stress-state that result from the new loading must then be imposed.
The strains which occur are measured and the settlement calculated.

For a complete description of this method, see Lambe (1964).

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.24
Topic 2C: Settlement of Shallow Footings

[Link] Davis and Poulos method

This method represents the first truly three dimensional method of analysis. The
necessary assumption is that the solid behaves as an ideal two phase material
consisting of a linearly elastic solid skeleton and an incompressible fluid filling the
voids. In this context, the fluid in the voids acts as a damping agent and does not
affect the ultimate behaviour of the elastic skeleton.

The 3-D field stresses are applied to a representative soil sample in a controlled
manner for both drained and undrained conditions. From elastic theory E’ and Eu
are determined from the laboratory test results. Elastic analyses are performed
using both the drained and undrained elastic parameters. The consolidation
settlement is the difference between those two values of settlement which have
been calculated from elasticity (Davis & Poulos, 1963, Davis & Poulos, 1968).

This method of course raises a most useful tool for the estimation of immediate and
total settlement for clays. If the porewater is regarded as simply a damping agent
and we can conduct laboratory (or indeed field) tests to give the undrained and
drained elastic parameters then immediate settlement for say a rectangular
foundation placed on an infinite soil layer is given by

1 𝜈
𝜌 𝑞𝐵 𝐼
𝐸

The total settlement for the same situation is given by

1 𝜈′
𝜌 𝑞𝐵 𝐼
𝐸′
It would appear then that the consolidation settlement would be given by

𝜌 𝜌 𝜌
Note: This method could only be regarded as approximate because the
consolidation component is assumed to be elastic which is not strictly true.
Therefore, this method is likely to severely underestimate c.

3.3 Secondary consolidation

This section discusses the possible mechanisms of secondary consolidation, and

presents the accepted method for computing secondary consolidation. This form of
settlement is relevant particularly to organic soils and highly compressible inorganic
clays. It is of little importance in overconsolidated clays.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.25
Topic 2C: Settlement of Shallow Footings

3.3.1 What is secondary consolidation?

This settlement occurs once primary consolidation is completed. Primary

consolidation is the process of transferring all excess pressure from the pore fluid
to the soil structure (i.e. from excess pore pressure to effective stress). Primary
consolidation occurs, therefore, as a result of changes in effective stress (or pore
pressure). Once the primary consolidation is complete, there can be no further
changes in effective stress.

Secondary consolidation, or secondary compression, as it is also referred to,

therefore occurs at a constant effective stress level and under no excess pore
pressure. Even under these conditions, further time-dependent volume changes
occur in the soil. This phenomenon is not yet well understood, however, it appears
to behave as a creep phenomenon i.e., a viscoelastic or viscoplastic phenomenon.
The most comprehensive treatment of secondary compression is by Mesri (1973).

Some researchers contend that the theory that secondary consolidation occurs at
zero excess pore pressure or constant effective stress is incorrect. Present evidence
seems to indicate that the primary/secondary division is purely artificial and arises
largely as a result of traditional laboratory testing techniques. The relationship
between "primary" and "secondary" appears to be very much influenced by the
thickness of the sample tested and attempts have recently been made to present a
unified theory which eliminates the distinction entirely. Their argument is essentially
that since water flows from the soil during secondary consolidation there must be a
small hydraulic gradient, and therefore small excess pore water pressures. This then
eliminates the distinction between secondary and primary consolidation.

3.3.2 Calculation of secondary consolidation

Once the primary consolidation is complete, the secondary compression continues

with a near linear reduction of void ratio with logarithm of time. The secondary
compression index is defined as follows:

Δ𝑒 Δ𝑒
𝐶 𝑡
𝑙𝑜𝑔𝑡 𝑙𝑜𝑔𝑡 log
𝑡

The settlement due to secondary consolidation can be calculated as:

𝐶 𝐻 𝑡
𝜌 log
1 𝑒 𝑡

As the amount of secondary settlement is a function of time, the total secondary

consolidation during the design life of the structure is calculated by substituting the
design life for t2. Given the uncertainty often associated with determining the rate

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.26
Topic 2C: Settlement of Shallow Footings

of primary consolidation, the time t1 must be chosen with due regard for the
differences between laboratory and field rates of consolidation.

3.3.3 Typical secondary compression index values

Mesri and Godlewski (1977) have compiled the ratio of secondary compression (C)
to compression index (Cc) for a number of soils as:

For inorganic clays and silts: 0.04 0.01

For organic clays and silts: 0.05 0.01

For peats: 0.075 0.01

4. Corrections to total settlement

The total settlement of a footing (𝜌 is determined by summing the immediate 𝜌 ,
primary (𝜌 and secondary consolidation (𝜌 components.

𝜌 𝜌 𝜌 𝜌

The above components have been computed with the assumptions that the loaded
area is flexible, and imposes a uniform pressure, and that the loaded area is at the
surface of the soil. These conditions are often not met, and corrections must be
applied to account for footing rigidity, and footings being buried.

4.1 Rigidity Correction

Settlement calculations are generally based on the assumption that a uniform load
exists over the foundation for all stages of settlement. This may be nearly true for
earth embankments, dams and tanks. However, it is not true for non-flexible bases
(i.e. those which are rigid or semi-rigid), such as rigid concrete footings or in the
semi-rigid class, compacted road bases.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.27
Topic 2C: Settlement of Shallow Footings

Q Q
q= q=
C O N TAC T  R2 2R (R 2 -a 2 )
STRE SS

SET TL EM EN T
PRO FILE

(a) (b)

For a flexible base in clay, the settlement profile would be non-uniform and pressure
would be uniform as shown in (a) above, whereas a rigid base would apply non-
uniform pressure and experience uniform settlement like (b) above.

Depending on the method of analysis it may be necessary to apply correction factors

for rigid bases. Sometimes this can be done by using the appropriate stress
distribution for bases. However, if the distribution for a flexible base is used, a
correction factor must be applied. This correction factor is known as the rigidity factor
(IR). Bowles (1997) suggested that the settlement of the rigid foundation can be
estimated as 93% of the settlement at the centre of a flexible footing.

The settlement profiles and contact pressures for flexible and rigid footings on sand
are quite different from those shown for clay in the above diagrams. This is because
the modulus of sand generally increases with depth, whereas that for clay tends to
be uniform, and because the lack of lateral confinement at the edge of the footing
tends to push the surface sand outward, reducing edge stresses for both types of
footing. Therefore, in the absence of any other information, the assumed rigidity
correction can be applied to footings on sand as well.

4.2 Depth Correction

The elastic solutions derived previously are invalid if foundation loads are applied at
some depth below the ground surface. Fox (1948) established the depth correction
factor (If) which showed that the calculated elastic settlement should be reduced by
a factor of between 0.5 and 1.0. The depth correction factors (If) for different values
of D/B, L/B and Poisson’s ratios () can be found in Das & Sivakugan (2019) (refer
to Figure 9.5 in 9ed.).

5. Time rate of settlement

In addition to calculating the amount of settlement, which a structure might
experience, it is often necessary to estimate how long the settlement process or at
least part of it will take.
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.28
Topic 2C: Settlement of Shallow Footings

An obvious example of this need is in situations when a clay site is to be preloaded

to bring about settlement prior to construction of a building. Since immediate
settlement is instantaneous, the consolidation component of settlement will be
considered in this section.

5.1 Principle of Estimation

Just as classic one-dimensional consolidation theory was the basis for estimating
consolidation settlements, this theory also provides the basis for estimating the rate
of settlement. In this theory, which you have learned in CIV3247: Geomechanics-II
unit, the essential time relationship is given by the expression below.

𝐶 𝑡
𝑇
𝐻
Where, T = time constant for the particular initial porewater pressure distribution
(dimensionless); Cv = coefficient of consolidation (m2/year); t = real time (yr); and
H = maximum length of the drainage path (m).

5.1.1 Excess pore water pressure distribution

In order to calculate the times of consolidation, it is necessary first to find the stress
distribution produced in the soil due to surface loading. With relatively thin
consolidating layers and wide foundations, this distribution can be approximated as
rectangular, whereas with thick soil formations and/or small loaded areas, a
triangular approximation may be more appropriate. Each case must be judged
individually.

Clay
Clay Approx.
triangular
Sand Approx.
rectangular

Sand

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.29
Topic 2C: Settlement of Shallow Footings

5.1.2 Time constant (T)

Standard time factor tables are available, which list values for T depending on the
drainage conditions and the initial excess pore water pressure distributions. This
diagram shows three such initial pore water pressure distributions. The following
table lists time constants T for values of U from 10 to 90% of consolidation
completed for the three cases. For difficult stress distributions, T may be found by
analytical methods if necessary i.e., finite difference techniques may be used.

Initial excess, PWP distr.

TOP
PERMEABLE

BOTTOM
PERMEABLE

Degree of consolidation (U%) Time Factor (T)

Condition (1) Condition (2) Condition (3)
10 0.008 0.047 0.003
20 0.031 0.100 0.009
30 0.071 0.158 0.024
40 0.126 0.221 0.048
50 0.197 0.294 0.092
60 0.287 0.383 0.160
70 0.403 0.500 0.271
80 0.567 0.665 0.440
90 0.848 0.940 0.720

5.1.3 The coefficient of consolidation

The value of cv is determined from the Oedometer test for the appropriate stress
range of the problem.

5.1.4 The maximum drainage path

As shown in the figure below, for a layer of thickness z, which is double drained (a),
H = z/2, but where only a single drainage layer exists (b), H = z.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.30
Topic 2C: Settlement of Shallow Footings

Sand Sand

Soft Soft
z z
Clay Clay

Sand Impervious Rock

H = z/ 2 H =z
(a) (b)

5.1.5 Time-settlement curves

Knowing T, Cv, H and t1, the percentage consolidation (U), can be computed, and a
time-settlement curve can be derived for the particular consolidation settlement. At
any time t, the settlement of a foundation is given by

𝜌 𝜌 𝑈𝜌
5.2 Problems in Practice

The previous section presents a picture which seems to indicate that the estimation
of rates of settlement is fairly simple. Unfortunately, in practice, the actual situation
is much more complicated.

5.2.1 Influence of radial drainage

In situations where the loaded area is relatively wide in comparison with the
consolidating layer the assumption made above concerning one-dimensional
consolidation theory is probably reasonable. However when the width of the loaded
area is about the same or smaller than the consolidating layer, then it is likely that
considerable radial drainage will take place. Clearly the magnitude of the horizontal
permeability will have a significant influence on this form of drainage. However, as
soil layers including the more permeable are laid down in horizontal layers, it is likely
that in many situations the horizontal permeability will be greater than the vertical.
It follows that the influence of radial drainage will be significant in most situations,
with the exception of those involving very wide loaded areas on relatively thin
formations of clay.

In order to make better estimates of time rates of settlement when radial as well as
vertical drainage occurs it is necessary to undertake much more complex analyses.
This may be in the form of three-dimensional theory or by use of an approximate
solution involving the superposition of vertical and radial drainage. Either way these
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.31
Topic 2C: Settlement of Shallow Footings

solutions are difficult and require a reasonable estimation of the respective

coefficients of consolidations, cv and ch. These parameters can be measured by
conducting the appropriate tests on appropriately oriented samples, and preferably
of sufficient size to overcome the second problem discussed below.

Another method which could be useful in establishing rates of settlement when

radial drainage is involved is to carry out large scale field loading tests using
instrumented trial embankments. The overall effect and character of the settlement
rate will be reflected in the performance of the trial embankment. By using one-
dimensional theory to back-calculate an overall cv to include the influence of radial
drainage, a reasonably representative coefficient of consolidation may be
determined for the site.

5.2.2 Influence of fabric

The other major influence on the actual rate of settlement is the presence of fabric
within a soil mass (Rowe, 1972). Fabric includes all the many forms of non-
homogeneities which are common in most soils: joints, fissures, fractures,
laminations, lenses and root fibres. These features tend to be much more permeable
than the intact soil and their presence tends to accelerate rates of settlement.

The problem arises because the sample of soil contained in an Oedometer test is
only very small. Therefore, even though a given layer of soil may contain much fabric
there is usually little of this fabric in the test sample. It follows that the results
obtained from such tests will indicate coefficients of consolidation considerably
lower than for the soil mass. A typical value for cv for an intact clay is around 1
m2/year. Depending on the fabric
10 6
Piezom eter readings close content for a soil, cv for the soil mass
to highly fissured rock
From full scale piezom eter
can be well in excess of 100 m2/
Coefficient of consolidation , cv - m2 / year

5
10 and settlement readings year, demonstrating the dramatic
From sm all scale influence of fabric.
laboratory testing.
10 4
This figure on the left shows cv
c lo s e ly sp
a c e d fi ss u values determined or inferred for a
re s
10 3 stiff fissured clay, and serves to
illustrate the effect of fabric. The
methods used to overcome this
10 2 problem generally involve attempts
we ll sp ac ed fis su
to obtain samples for testing so that
re s a representative component of fabric
10
is included. An appropriate size is
one where the sample tested is
0
1 100 200 300 400
several times the spacing of the
Effective stress - kPa
fabric. For example, if fissures are
spaced at about 100 mm then the
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.32
Topic 2C: Settlement of Shallow Footings

sample required would need to be in excess of about 300 mm (about the size of a
"Rowe" consolidation cell), perhaps somewhat more. Clearly there is a limit to the
size of sample that can be retrieved for laboratory testing. Alternatively, larger scale
field testing may have to be carried out. Once again for the larger projects where
more precise rate estimations are required, it may be economically feasible to
construct large instrumented trial embankments from which a realistic cv can be
back-calculated.

6. Swelling and groundwater changes

The previous sections have dealt exclusively with settlement due to foundation
loading. Removal of soil loading in an excavation will cause swelling, and
fluctuations in the groundwater level may also cause settlement or swelling. These
matters are discussed here.

6.1 Swelling of Excavations

This is the opposite of settlement and can occur after opening an excavation i.e.,
the removal of overburden. The swelling involved can be calculated by the normal
methods, but with a negative stress or porewater pressure change. Swelling effects
will be restricted to excavations in clay.

In general, swelling is insignificant in the case of shallow foundations and can

usually be neglected. It must however be taken into account for deep excavations
where significant swelling may take place over a period of time.

The swelling is associated with an increased moisture content and void ratio, and
therefore with a loss of soil strength.

6.2 Ground Water Table Changes

Should the ground water table change by any significant amount during the lifetime
of a building changes of effective stress in the soil will result. These stress changes
may bring about additional consolidation or swelling.

For example, should the ground water table fall the effective stress in the soil will
increase as the soil particles have to support the full weight rather than the buoyant
weight of the overlying soil. This will result in additional consolidation. Pumping
water from aquifers below the consolidating soil, and thereby lowering the porewater
pressures has the same effect. Such effects are most vividly seen in Venice,
Bangkok and Mexico City.

Mexico City has also suffered this settlement, as much as 7m total and 2m
differential in parts of the city. Bangkok is apparently settling at a rate of about
10mm/year (Zeevaert, 1953).

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.33
Topic 2C: Settlement of Shallow Footings

7. Overall summary
When the elastic and consolidation properties of a soil can be derived by appropriate
laboratory and/or in-situ tests, the total amount of settlement is given by

𝜌 𝐼 𝐼 𝜌 𝜇𝜌 𝜌

Note: Some elastic solutions may include the rigidity and depth factors in it and one
should not apply again to estimate the final settlement. If the stress calculation is
based on the assumption of a flexible surface footing then the above corrections are
applicable.

8. Limiting values of total settlement

The routine limits for the maximum total settlement for buildings (f) are summarised
in the following table (O’Brien, 2012).

Foundation type and soil Maximum total settlement

Pad footing in clay 65 mm

Pad footing in sand 40 mm
Raft in clay 100 mm
Raft in sand 65 mm

Terzaghi & Peck (1948) suggested an allowable settlement value of 25 mm for

isolated pad footing and 50 mm for raft foundation.

9. Settlement based on field tests

As discussed previously settlements in cohesive soils comprise both an elastic
component which occurs immediately and a (generally larger) consolidation
component which occurs over an extended time.

To describe the settlement of these soils, both undrained elastic parameters, Eu

and u are required, as well as consolidation parameters such as mv or Cc.
Cohesive soils are generally easily recovered, and can therefore be tested in the
laboratory, however, there may be advantages in testing these soils in-situ to
minimize the effects of disturbance that occur with sample retrieval.

In non-cohesive soils, all settlements occur almost immediately. Nevertheless, just

as for the cohesive soils, part of this settlement is elastic and the other part is
consolidation, albeit instantaneous, which is neither elastic nor time dependent.
This second part can be treated in exactly the same way as the elastic settlement

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.34
Topic 2C: Settlement of Shallow Footings

by the application of the appropriate values of E and . It follows that for the
settlement of non-cohesive soils, the elastic methods described previously should
be used if the elastic parameters can be established. It is almost impossible to
retrieve samples of non-cohesive soils for testing in the laboratory and so in-situ
methods must be used to estimate the appropriate elastic settlements. The correct
estimation of these parameters is particularly important as settlement is generally
the controlling factor for foundations on sands.

A range of rather approximate methods of settlement evaluation have evolved using

extrapolation from field tests and semi-empirical methods. These are described in
the following sections.

9.1 Settlement based on SPT

Meyerhof (1956) presented a correlation between the net bearing pressure and SPT
blow count (N60) for 25 mm of maximum settlement of shallow footing in sand.
Bowles (1977) proposed a modified form of the correlation to incorporate the depth
factor (Fd) and elastic settlement as:

𝑞 𝑘𝑃𝑎 𝐹 (for B ≤ 1.22 m)

.

.
𝑞 𝑘𝑃𝑎 𝐹 (for B ≥ 1.22 m)
.

Where, Fd=depth factor=1+0.33(D/B); B=width of footing in meters; Se=settlement

in mm; D=depth of footing in meters; N60=average SPT blow counts over 2B depth
below footing.

Burland & Burbidge (1985) proposed a method of calculating the elastic settlement
of shallow footing in sandy soils based on the SPT N60. Details of this method can
be found in Das & Sivakugan (2019).

9.2 Settlement based on CPT

Schmertmann et al. (1978) proposed a semi-empirical method of calculating the

elastic settlement of shallow footing in sand based on the cone resistance (qc) as:

𝐼
𝑆 𝐶𝐶 𝑞 𝑞 Δ𝑧
𝐸

Where, C1=depth embedment factor=1 0.5 0.05

C2=creep factor=1+0.2log(t/0.1), t=time in years

𝑞 =stress below footing; q='D=effective stress at the founding depth
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.35
Topic 2C: Settlement of Shallow Footings

Es=modulus of elasticity of soil

Iz=strain influence factor (refer to figure below for square/circular and strip footing
cases)

Schmertmann et al. (1978) suggested Es as:

Es=2.5qc (for square footing)
Es=3.5qc (for L/B≥10) (for strip footing)
𝐸 2.5𝑞 1 0.4𝑙𝑜𝑔 (Terzaghi et al., 1996)
By summing up the settlement for each layer of soil within the entire depth of
influence below the footing, the total elastic settlement can be computed. A step by
step procedure can be found in Das & Sivakugan (2019) and Bowles (1997). A slight
modification of the strain influence diagram of Schmertmann et al. (1978) was made
by Terzaghi et al. (1996).

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.36
Topic 2C: Settlement of Shallow Footings

9.3 Settlement based on Pressuremeter test

The pressuremeter stress-strain results over the depth of influence zone below the
footing (refer to CPT) can be used to calculate the elastic settlement in sand based
on a method proposed by Briaud (2007). This method will not be explored here.

9.4 Settlement based on Plate Bearing Tests

As full-scale load tests are generally either impractical or excessively expensive to

conduct, model tests are sometimes conducted using a 300mm wide steel plate.
These tests act as a scale model of the foundation to be constructed, and despite
their reduced cost, are still relatively expensive. Plate bearing tests are seldom used
in practice.

The major problem with the plate bearing test is that the zone of influence of the
300mm plate is substantially less than that of the full-scale footing - at approximately
4B, the zone of influence is 1.2m. By comparison, for a 3m wide footing, the zone
of influence is about 12m. If soil conditions improve (or deteriorate) with depth, or
the soil is stratified, the results of the plate bearing test can be very misleading.

Despite the difficulties associated with the plate bearing test, for soils with large
boulders and cobbles (even fill materials) this method may be the only reasonable
method of predicting settlements.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.37
Topic 2C: Settlement of Shallow Footings

Terzaghi and Peck (1967) give the following relationship for the interpretation of the
test.

2𝐵
𝑆 𝑆
𝐵 0.3
Where, S= settlement of foundation; S1=settlement of 0.30 m square plate; B= width
of foundation.

It is noted that this method is not too satisfactory for loose and organic soils. The
plate bearing test is described in more detail in Bowles (1997).

10. Effect of vibrations

Vibrations transferred to the soil by machinery (compressors etc.) may lead to very
large settlements (and differential settlements).

The theory will not be considered now but suffice it to say that where such
occurrences may occur special care should be taken in the design of foundations.

A general design guide to minimize effects is as follows.

 Design foundations to absorb vibrations (i.e. the bigger the better).

 Use a low value of design pressure, qd.
 Increase low density soils by means of ground improvement techniques
e.g. vibro compaction, dynamic compaction, grouting etc. These soils
are generally the ones which cause major problems unless densified.

In earthquake zones it must be remembered that settlement could be critical. For

loose soils, liquefaction may occur with possibly disastrous results.

11. Differential settlements

In the preceding sections, the settlement of individual footings has been
investigated. These have been assumed to settle uniformly. If soil conditions are
variable below a footing, it may not settle uniformly.

Similarly, individual footings that support a structure may settle by different amounts
for a variety of reasons. These differential settlements can cause distress in the
structure if they are not taken into account in the initial structural design. This section
covers the reasons for differential settlement, and the approach used for its
estimation.

Often if the whole structure settles uniformly the structure itself may suffer no
distress. However, differential settlements can be extremely troublesome.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.38
Topic 2C: Settlement of Shallow Footings

 If differential settlement is excessive there may be load transfer to other

foundation areas enough to cause bearing capacity failure.
 Structural distress (cracking) and possible failure of the superstructure.

In addition to an assessment of total settlement it is therefore vitally important that

the likely differential settlement is evaluated.

For differential settlement, a series of empirical guidelines have been proposed so

that excessive differential settlement damage does not occur. Obviously these
depend on the form of construction adopted and it is difficult to apply definite rules.
It will also depend on equipment within the construction.

11.1 Settlement sensitive structures

 Masonry structures (rigid and likely to crack)

 Reinforced concrete structures (rigid and likely to crack but not as critical)
 Concrete structures with glass curtain walling (rigid glass cracking)
 Continuous span bridges (excessive settlement of one support)
 Structures carrying long high tolerance machinery (for precision)
 Oil tank walls (to stop buckling and jamming of roof)

11.2 Settlement tolerant structures

 Timber framed and clad structures

 Steel framed and clad structures (but not too high)
 Simply supported bridges
 Oil tank floors (care!)

11.3 Reasons for differential settlement

Differential settlement can occur for a whole variety of reasons. However, these can
be conveniently grouped as those due to discernible differences in boundary
conditions, and inherent differential settlement. These are outlined in the following
sections.

11.3.1 Discernible differences in boundary conditions

Examples of these are as follows:

 Variations in soil strata such as lenses of sand in clay strata or vice versa.
Irregular bedrock surface.
 Variations in foundation loading such as towers with low side wings and
machinery foundations.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.39
Topic 2C: Settlement of Shallow Footings

 Variations in foundations e.g. partly piled partly earth supported or partly

in excavation.
 Effect of large flexible loaded areas which leads to characteristic bowl
shaped settlement curve.
 Difference in time of construction e.g. extensions.
 Variation in site conditions - such as partly old demolished buildings
which will have preconsolidated parts of the soil.

In all cases, it is possible (indeed necessary) to make an engineering estimate of

the likely differential settlement effects caused by the different boundary conditions.

11.3.2 Differential settlement inherent in total settlement

Another important source of differential settlement is due to the inherent

simplifications that are made during the design process. We generally assume
uniform, or nicely stratified soil conditions, when in reality, the soil varies greatly
across a site, even in a short distance. In a practical sense, it is not possible to
detect all these variations in soil characteristics which can affect foundation
performance. Even if they could be detected it would be very difficult to analyse such
a large number of variables within the usual technical and economic constraints of
a project.

If a reasonable investigation has been conducted, and representative soil

parameters chosen, the performance predicted by our simplified models should be
close to the average performance of the actual foundations. However, the soil
variations will cause some variability around the average performance which cannot
easily be quantified.

The following table shows some typical differential settlement values for shallow
foundations in terms of maximum settlement values (O’Brien, 2012).

Foundation type and soil Typical differential settlement

Pad footing in clay < 0.66 max, if max ≤50 mm

< 0.5 max, if max >200 mm

Pad footing in sand < 0.75 max

Raft in clay < 0.33 max

The approximation of differential settlement values based on the maximum

estimated total settlement value could help the designer to check for the distortion
Department of Civil Engineering, Monash University
Revised: 5/2020
CIV4249: Foundation Engineering 2C.40
Topic 2C: Settlement of Shallow Footings

limit or vice-versa. Bjerrum (1963) recommended the following limiting angular

distortions (/ L).

 / L = 1/150. Safe limit for flexible brick walls with L/H>4. Danger of
structural damage to most buildings. Cracking of panels and brick walls.
 / L = 1/250. Visible tilting of high rigid buildings.
 / L = 1/300. First cracking of panel walls
 / L = 1/500. Safe limit for buildings where cracking is not permissible.
 / L = 1/600. Danger limit for frames with diagonals.
Obviously all these are dependent on the structure. Therefore, each structure should
be designed based on its own merits and code specified values.

11.4 Methods of Reducing Differential Settlement

If a structure cannot accommodate the calculated amount of differential settlement

it may be necessary to adopt one of the following situations.

 Provision of a rigid raft with thick slab or deep beams.

 Provision of deep basement
 Transfer of load to deeper less compressible soils by piles or piers etc.
 Pre-treatment of the site with appropriate methods e.g. vibrocompaction
dynamic compaction, preloading or grouting
 Provision of jacking points

12. Other causes of ground movement

Certain other forms of ground movement can occur, which are independent of
ground loading. These include:

 Frost Heave
 Hillside creep
 Mining subsidence
 Swallow holes
 Shock loading - e.g. earthquakes nearby collisions!
 Shrinkage of soils and swelling due to seasonal moisture and
temperature changes
 As for the previous, but due to vegetation etc.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 2C.41
Topic 2C: Settlement of Shallow Footings

 Change of structure of soils (especially clays) by base exchange,

possibly by pollution.
 Drawdown of water table or changes in porewater pressures

13. References
Bowles, J.E. (1997). Foundation analysis and design. 5th ed. McGraw-Hill.

Bjerrum, L. (1963). Allowable settlement of structures. Proc. European Conf. on Soil

Mech. & Foundation Engg. Wiesbaden, Germany, Vol. III, 135-139.

Briaud, J.L. (2007). Spread footing in sand: Load settlement approach. J. Geotech.
Geoenv. Engg., ASCE, 133(8), 905-920.

Das, B.M. & Sivakugan, N. (2019). Principles of Foundation engineering. 5ed.

Cengage publishing, USA.

Davis, E.H. & Poulos, H.G. (1968). The use of elastic theory for settlement prediction
under three dimensional condition. Geotechnique, Vol. 18, 1968.

O’Brien, A. S. (2012). Foundation types and conceptual design principles. Chapter

52, ICE Manual of Geotechnical Engineering. ICE publishing, Vol III, 733-764.

Lambe, T.W. (1964). Method of estimating settlement. ASCE. SM5, Sept. 1964.

Mesri, G. (1973). Coefficient of secondary compression. J. Soil Mech. and Found.

Div. ASCE, Vol. 99, No. SM1: 122 - 137.

Mesri, G. & Castro, A. (1986). C/Cc concept and Ko during secondary compression.
J. Geotech Engg. ASCE, Vol. 113, No. 3 : 230 - 247.

Parry, R.H.G. (1968). Field and laboratory behaviour of a lightly overconsolidated

clay. Geotechnique June 1968.

Rowe, P.W. (1972). The relevance of soil fabric to site investigation practice.
Geotechnique June 1972

Skempton, A.W. & Bjerrum, L. (1957). S contribution to the settlement analysis of

foundations on clay. Geotechnique, December 1957.

Terzaghi, K., Peck, R.B. & Mesri, G. (1996). Soil Mechanics in Engineering Practice.
John Wiley & Sons, New York.

Department of Civil Engineering, Monash University

Revised: 5/2020
CIV4249: Foundation Engineering 3A.1
Topic 3A: Axial Capacity of Piles

Topic 3A: AXIAL CAPACITY OF PILES

TABLE OF CONTENTS

1. AXIAL CAPACITY OF SINGLE PILES 4

1.1 Static formulae 4

1.2 Pile driving formulae 4

1.3 Static load tests 5

1.4 Dynamic testing 5

2. GENERAL APPROACH 5
2.1 Ultimate base resistance 6

2.2 Ultimate shaft resistance 7

2.3 Ultimate axial pile resistance 8

2.4 Difficulties in application 8

2.4.1 Soil disturbance 8
2.4.2 Scale effects 9
2.4.3 Loading rate effects 10
2.4.4 Porewater pressure effects 10
2.4.5 Friction angle determination 11
2.4.6 Lateral pressure 12
2.4.7 Bearing failure 12

2.5 Closing remarks 12

3. SINGLE PILES IN CLAY 13

3.1 Total stress approach (undrained condition) 13
3.1.1 The generalized pile capacity equation 13
3.1.2 The Nc Parameter 14
3.1.3 Nc in layered soils 14
3.1.4 The value of cb 15
3.1.5 The value of cs 16

3.2 Effective stress approach (ESA) for shaft resistance in clay 19

3.2.1 The generalized pile capacity equation in clay based on ESA 19
3.2.2 Base resistance of clay based on ESA 21

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.2
Topic 3A: Axial Capacity of Piles

3.2.3 Closing remarks 22

3.3 The lambda method 22

4. SINGLE PILES IN GRANULAR SOILS 23

4.1 The generalized pile capacity equation 24
4.1.1 The Nq parameter 24
4.1.2 Meyerhof’s method 24
4.1.3 Berezantzev’s method 26

4.2 The value of P'ob 27

4.2.1 Unlimited depth method - Meyerhof 28
4.2.2 Limited depth method – Vesic (1967) 28

4.3 Theory of expansion of cavity-Vesic’s method 29

4.4 Kulhawy’s Method 30

4.5 The value of P'os 30

4.6 The parameters K and  31

4.6.1 The lateral pressure coefficient K 31
4.6.2 The pile-soil friction angle,  31
4.6.3 The combined parameter K tan  32

4.7 Direct correlations with insitu SPT/CPT data 34

4.7.1 Shaft and end bearing of piles based on SPT 35
4.7.2 Shaft and end bearing for piles based on CPT 38

5. UNDERLYING WEAK LAYERS 40

6. SINGLE PILES IN ROCK 41

6.1 End bearing 41
6.1.1 Goodman’s method 41
6.1.2 Zhang’s method 42

6.2 Socket resistance 42

6.2.1 Empirical methods 42
6.2.2 Kulhawy’s method 43
6.2.3 Other notable methods 43
6.2.4 Pells method 44
6.2.5 Monash method 45

7. AXIAL CAPACITY OF PILE GROUPS 46

8. PILE GROUPS IN CLAY 47

8.1 Free standing pile groups 47
8.1.1 Feld Rule 48

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.3
Topic 3A: Axial Capacity of Piles

8.1.2 Converse-Labarre Formula 48

8.1.3 Block Failure 49

8.2 Capped pile groups 51

9. PILE GROUPS IN GRANULAR SOILS 52

10. REFERENCES 54

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.4
Topic 3A: Axial Capacity of Piles

1. Axial Capacity of Single Piles

There are a number of methods, which can be used to establish the axial load
capacity of single piles. These include static formulae used by the designer in
advance, static load testing which is performed after installation, dynamic
formulae which are applied during installation of driven piles and dynamic testing
which may be performed during or after installation.

This chapter deals exclusively with the use of static formulae, which is the basis
on which design usually rests. Occasionally, and particularly for very large
projects, static formulae will be correlated against static or dynamic load testing
as part of the design process, however, in general, such testing is only conducted
during the contract period. Such confirmation is very important, as soil and rock
conditions on site may vary considerably from those established in the (limited)
site investigation. In addition, static formulae are based on approximate models
of soil and pile behaviour that must be verified by actual load testing. Some form
of load testing is carried out on most piling projects. The various load test methods
will be covered elsewhere.

1.1 Static formulae

This is the application of geomechanics theory to predict the capacity or deflection

performance of a single pile prior to installation. These theories are applied using
the knowledge of the stratigraphy and geotechnical properties determined from
the site investigation.

1.2 Pile driving formulae

Pile driving formulae are used to confirm that the capacity of driven piles is in
accordance with the design and/or contractual requirements. These formulae
relate pile capacity to the hammer energy. The most common formula used in
Australia is the Hiley Formula. In the U.S., the Engineering News Formula is used.
High factors of safety should be used because of the following (incorrect)
assumptions common to all methods:

 the pile is a rigid body (not true, it is elastic)

 the soil resistance is all static (not true, there are large dynamic forces during
pile driving)
 the efficiency of the hammer drop is known (can vary greatly)

In order to apply the dynamic formulae, workmen have to “take the set” on the
pile.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.5
Topic 3A: Axial Capacity of Piles

1.3 Static load tests

Static load tests are performed either to establish the ultimate capacity of the pile,
or to ensure that the pile has a sufficient factor of safety (1.5 to 2.0 is common)
and a satisfactory performance at working load.

In a static load test, the load is gradually applied according to a predetermined

schedule by a hydraulic jack. The jack reacts against a large beam which either
supports a physical mass (kentledge) or resists anchors that are socketed into
the soil or (usually) rock. Load tests may be conducted to 3000 tonnes or more.

Static load testing is generally considered the definitive assessment of pile

capacity, although if poorly designed and conducted, results could be questioned.
This type of testing is expensive, and time consuming (3 days to set up, conduct
and dismantle).

1.4 Dynamic testing

Dynamic pile testing is a relatively recent technology (1980) based on

measurement of stress waves in piles during the driving process. The hammer
impact generates a downward travelling stress-wave; soil resistance generates
upward travelling stress-waves. These stress-waves can be measured by
accelerometers and strain gauges and separated into the two directions. By
evaluating the magnitude and timing of the upward travelling stress-waves, the
pile capacity can be established, as well as the distribution of soil resistance. It
should be noted, however, that the soil resistance comprises both static and
dynamic components, and it is necessary to establish the static component only.
This is achieved by numerical simulation of the driving process using a matching
technique on the actually measured data. On the basis of this analysis, a static
load-settlement curve can be predicted.

It should be noted that this technique is often used for construction control, as the
energy transfer efficiency of the pile driving hammer, the maximum compressive
and tensile stresses in the pile, and the location and severity of any structural
damage in the pile can be established.

2. General approach

The resistance of a pile has two components: the side resistance and the base
resistance. It is generally assumed in design that these components can be
computed independently and then summed.

The expression relating pile load at failure (Pu) to ultimate soil resistance is
therefore given by

Pu  Qb  Qs  Wp

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.6
Topic 3A: Axial Capacity of Piles

Where, Qs= the ultimate shaft resistance; Qb= the ultimate base resistance; and
Wp= the weight of the pile.
2.1 Ultimate base resistance

The ultimate base resistance can be estimated from the basic Terzaghi’s bearing
capacity equation as:

 B 
Qb  Ab c b N c  P 'ob (N q  1)  N  Pob 
 2 
Where, Ab= the area of the base; cb= the representative cohesion at the base of
the pile; Nc, Nq and N = bearing capacity factors; Pob and P'ob=the total and
effective vertical overburden stresses at the base of the pile; = the
representative effective unit weight at the pile base; and B= the width of the pile
base.

If the weight of the pile is included, the expression for the ultimate base resistance
becomes

 B 
Qb  Ab c b N c  P 'ob (N q  1)  N  Pob   Wp
 2 
However, the weight of the pile is similar to the weight of the soil displaced or
replaced by the installation of the pile, i.e.

Wp  Ab Pob

and further, for a pile length, L, being very much greater than its width, B, it is
reasonable to assume that

B
Wp  Ab N   Ab Pob
2
Similarly, for  > 0, Nq  Nq - 1.

Therefore, the ultimate base resistance of a pile can be approximated as

Qb = Ab [c b N c + P 'ob N q ] This equation is valid for c-soil.

Ultimate end bearing of pile in undrained clay (u=0): 𝑄 𝐴 𝑐 𝑁

Where, Nc=bearing capacity factore, a value of 9 is usually considered
(Skempton, 1951). If the base soil is disturbed or emoulded then a reduced value
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.7
Topic 3A: Axial Capacity of Piles

of undrained shear shtrength is to be used. If a pile just sits on the clay strata
then a reduced value of Nc=6 is to be used and Nc=9 at 3D (D=diameter of the
pile) penetration in the bearing layer (Bell & Robinson, 2012).

Ultimate end bearing of pile in sand (c’=0): 𝑄 𝐴 𝑃′ 𝑁

Where, Nq=bearing capacity factore; P’ob=effective vertical stress at toe level. Nq
is correlated with the friction angle (Berezantzev et al., 1961).

2.2 Ultimate shaft resistance

For a soil possessing both cohesion and friction, there will be two components of
shaft resistance - these are treated separately.

The cohesive component of the total shaft resistance (Qsc) will be a function of
many factors, but for simplicity, it is assumed that it is directly proportional to the
cohesion of the soil, i.e.

Qsc = As  c s

Where, As=total circumferential area of the pile shaft; cs=representative

(average) undrained cohesion on the pile shaft; = a correlation factor - effectively
a cohesion reduction factor.
For frictional soils, the effective overburden at any point on the pile shaft will be
P'os. At this same point, the effective horizontal stress will be KP'os, where K is
an earth pressure coefficient, which is the ratio of
. . .
. .
horizontal to vertical effective stresses. It follows that
.
. . . the effective stress normally to the side of the pile is
Pos .
also KP'os as illustrated in this schematic.
. . .
. .
.
.
. .
. The frictional component of the shaft resistance (Qsf)
Pile is this normal stress on the side of the pile multiplied
. . . by the tangent of the angle of friction () between the
KPos .
. .
soil and the pile material, i.e.
. . .
.

.. . Qsf = As KP 'os tan 

.
.
.
. . .
.
It follows that the total shaft resistance is given by the
sum of the cohesive and frictional components, i.e.

Qs = Qsc + Qsf = As [cs + KP'os tan] This equation is valid for c-soil.

Ultimate shaft resistance for undrained clay: 𝑄 𝐴 𝛼𝑐

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.8
Topic 3A: Axial Capacity of Piles

Where, As=area of the pile shaft; =adhesion factor; cs=average undrained shear
strength of the clay (cs=su).

Ultimate shaft resistance for sand: 𝑄 𝐴 𝐾𝑃′ 𝑡𝑎𝑛𝛿

Where, As=area of the pile shaft; =earth pressure coefficient which depends on
pile installation method; P’os=effective vertical stress; =pile wall-soil friction angle
which depends on pile material and soil type.

2.3 Ultimate axial pile resistance

Adding together the individual components of side and base resistance yields the
general expression for the total resistance of a pile loaded in axial compression.

Pu = Qb + Qs = Ab [c b N c + P 'ob N q ]  As [c s  KPos

 tan  ]

However, since the unit shaft resistance along the full length of the pile may vary
significantly because of differences in the types and properties of soils
encountered, the cross-sectional area of the pile and the effective overburden
pressures, it would be more correct to express this relationship as

L
Pu  Ab [c b N c  P 'ob N q ]   C [ c s  KP 'os tan  ] dL
0

Where C is the length of the pile perimeter [e.g., circular pile of diameter
D=C=πD; square pile of side B=C=4B].

2.4 Difficulties in application

Within the limitations of the assumptions discussed above, the above equations
are reasonably correct for the general case of any pile in any soil for either drained
or undrained conditions as may be appropriate. However, major difficulties arise
when the equations are applied. Some of the reasons for these difficulties are as
follows.

2.4.1 Soil disturbance

The aim of any site investigation and laboratory test program is to retrieve
samples with a minimum of disturbance, so that the in-situ strength can be
estimated as closely as possible.

However, the installation of a pile causes a great deal of disturbance. Driven piles
are forcibly inserted into the soil, resulting in soil being displaced down and out.
The stress levels in the soil increase, and with time, the strength of the clay could
be expected to increase accordingly. On the other hand, the soil is failed by the
insertion process, and a remoulded strength could be expected to apply.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.9
Topic 3A: Axial Capacity of Piles

If a pile is constructed by drilling, the same remoulding of soil at the wall of the
hole could be expected. However, as opposed to a driven pile, removal of the soil
by drilling will cause a reduction of stress state, and possible loss of strength.

In addition, the water used to assist the drilling can soften the walls of the hole
being drilled.

The laboratory-determined strength values may therefore be quite different from

the soil adjacent to an installed pile. The effect of construction processes on
soil strength should ideally be incorporated in the analysis.

2.4.2 Scale effects

The values of cohesion are usually established on relatively small samples of clay
(say 50 to 100 mm of diameter, and length to diameter ratio of 2). By contrast,
the volume of soil involved with failure of a pile under load is significantly greater.

If the soil is isotropic and homogeneous, this difference will not have any
consequence. However, soils in general contain significant fabric. In the case of
fissured clays, the structure is one of intact blocks surrounded by weaker
discontinuities. In such a case, the result of a test on an intact sample will
overestimate the mass strength of the soil (Rowe, 1972).

This figure shows the results of a number of strength tests conducted on fissured
London clay. The strength was determined by 5.5 mm diameter penetration tests,
38 and 98 mm diameter triaxial tests, and 865 mm diameter plate load tests. As
the figure shows, the strength measured by the small tests is very much greater
than the larger triaxial
Shear strength kN/m2
tests. These triaxial
tests in turn also show 0 100 200 300 400 500 600
greater strength than 865mm [Link] tests
obtained with large scale 38mm diam. triaxial Sand and gravel
98mm diam. triaxial
tests (i.e. the size of
Depth below G.L. ( metres )

Penetration tests
piles). The reason for
5 Soft brown clay
this is that with fissure
spacings in the range 5.5mm diam. penetration
Brown and grey
between about 100 and tests
mottled clay
250 mm, the smaller 10 38mm and 98mm
tests are unlikely to be diam. triaxials
Stiff grey clay
influenced by the weaker
fissures and therefore a
high strength is 865mm diam.
15
recorded. For the triaxial plate tests
tests, the fissures will
have some effect and
therefore the recorded
strength is reduced. For the large plate tests, the full influence of the fissures will
come into effect to produce an even lower strength.
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.10
Topic 3A: Axial Capacity of Piles

For this reason, for piles founded in fissured clays, it is common to introduce a
reduction factor for the cohesion acting at the pile base. The cohesion cb is often
replaced by cb, where  < 1. This reduction factor allows for the possibility that
bearing failure will be partly along fissures and partly in intact material.

By contrast, the shaft cohesion is not similarly reduced, as failure of the shaft is
locally at the shaft soil interface and does not allow for failure through fissures in
the body of the clay.

2.4.3 Loading rate effects

There is also a rate effect which causes a difference between the cohesion that
is derived from laboratory or in-situ tests and the cohesion that applies to a loaded
pile.

Soil investigation tests are normally conducted fairly quickly, and it is common
(although not universal) to base pile capacity calculations on undrained strength
parameters. By contrast, the dead loads and permanent live loads which often
account for the majority of load on a pile are gradually applied over several
months, if not years. This gradual loading allows the soil to consolidate and gain
strength under the imposed loading.

2.4.4 Porewater pressure effects

As noted previously, the installation of a driven pile causes soil to be displaced

downward and outward to accommodate the pile. The pile effectively applies a
large horizontal load to the soil. In clay soils, this load is initially balanced by very
high porewater pressures that can often exceed the overburden pressure. As the
vertical total
stress does 2.0
not change,
the effective
vertical stress 1.5
Average curve for sensitivea
may become marine clay
very small, u
1.0
and the soil vo Average curve for clays of
may lose low-medium sensitivity
strength 0.5
dramatically
whilst these
pore
pressures 0 10 20 30 40 50 60
r
remain a
elevated.
While this figure shows that the pore pressure drops very rapidly with distance
away from the pile, it must be remembered that it is the local pressure at the
interface between pile and soil which is critical for shaft resistance. Again, soil

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.11
Topic 3A: Axial Capacity of Piles

strengths determined from triaxial tests may bear little relevance to the strength
of clay soils during pile installation.

It should be noted that with time, the elevated pore pressures generated during
pile driving dissipate, and with this dissipation, consolidation of the soil occurs.
This consolidation is accompanied by strength increases in the clay. In practice
this is observed as an increase in resistance to driving after pile installation. Piles
that may have driven 50mm with each hammer blow during installation may only
move 0.1mm or less hours or days after installation. This effect is commonly
called pile setup or pile freeze (US). Pile set-up has been observed to continue
(at a decreasing rate) over months or years, as demonstrated by the field
research results shown below.

The set-up effect is also the main reason why pile load tests are preferably
conducted a minimum of 2 days (but preferably 7 days) after installation.

300 30

Bearing capacity in tons

Bearing capacity in kN

200 x 215mm conrete

250 25
(Gothenberg)
200 20

150 15
300 x 150mm tapered timber (Drammen)
100 10
300 x 125mm I-Beam 150mm (8 in) steel tube (San Francisco)
50 5
(Gothenberg)

0
1 5 10 50 100 500 1000
Time after driving in days

2.4.5 Friction angle determination

To determine the end bearing and shaft resistance components of pile capacity
in frictional soils, it is necessary to determine the soil friction angle (), which is a
measure of its strength. Although this value doe not appear in the above bearing
capacity equations, it is implied in the bearing capacity factors, Nc and Nq, and
in the pile-soil friction angle ().

Determination of the friction angle () for granular soils is very difficult, as it is not
possible to retrieve undisturbed samples of such materials. In general, it is
necessary to perform in-situ tests, and apply some empirical formulae to the tests
to infer a friction angle. This process is of itself fraught with danger and
complications, as boundary conditions in these tests cannot be controlled, and
the parameters which are important are often not well understood - see for
example the many and conflicting correction factors which are proposed for SPT
tests, or proposed material correlations for CPT tests.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.12
Topic 3A: Axial Capacity of Piles

Now add to this the effects of pile installation, which could be densification in the
case of driven piles, or stress-relief (loosening) in the case of drilled piles. The
friction angle of the sand is intimately related to the density (or relative density) of
the granular material i.e., how closely the particles are packed. It is evident that
the post-installation friction angles and in-situ friction angles will differ, and that
proper account of these differences must be made in the design process.

2.4.6 Lateral pressure

The value of K which applies to the frictional component of shaft resistance is a

very difficult parameter to establish for a given pile installation. It is hard enough
to establish the "at rest" value, Ko, for a soil, but to complicate it by the many
influences of pile installation is to make its measurement virtually impossible for
routine developments. As the value of K is very sensitive to soil density, it can
vary a great deal even over a short length of pile shaft.

2.4.7 Bearing failure

The bearing capacity factors, Nc and Nq, have been developed from approximate
theories by a number of different researchers. These theories have also been
modified empirically to make theory and practice match more closely. Depending
on which theory is adopted, different values of these parameters will be obtained.

The bearing capacity factors are principally dependent on the value of  adopted
(in itself a difficult choice). For low values of , the bearing capacity values are
generally insensitive, however, for high values of , Nc and Nq vary considerably,
and a correct choice is critical.

2.5 Closing remarks

The application of the generalized pile capacity equation is deceptively simple. It

has been shown that choice of appropriate parameters to apply to this equation
is critical, and will make the difference between an inappropriate and appropriate
capacity estimate.

One of the most critical aspects of pile design is proper consideration of the
influence of installation method on the strength of the soils that support the pile.
Driving processes which displace soil tend to increase soil strength and pile
capacity, whereas drilling processes which tend to remove soil tend to decrease
soil strength and pile capacity.

It would seem from the above that driving processes would in all cases be the
preferred method, however, choice of pile type is dependent on column loads,
soil type and stratigraphy, equipment availability, noise and vibration concerns,
historical precedent, labour considerations, and most of all price. The choice is
by no means straight forward.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.13
Topic 3A: Axial Capacity of Piles

It would be normal for most projects to check the design by means of one of the
other methods mentioned previously (e.g. static or dynamic test loading of
selected piles). This is particularly appropriate for bored piles for which there is
no driving record. In general, the more significant the job, the more sophisticated
and expensive the testing program justified. For small jobs, because of the
expense of such testing, it is more usual to adopt conservative estimates of the
variables used in the static formulae; driven piles may be checked for these jobs
using only driving formulae with high factors of safety.

3. Single piles in clay

In designing piles in clay it is possible to take two diametrically opposed

approaches: the total stress approach and the effective stress approach. Both
have their advantages and their limitations.

3.1 Total stress approach (undrained condition)

The conventional approach for piles installed in clay is to assume that the clay is
saturated and undrained conditions apply. This implies that the angle of friction
of the clay () is zero and the relevant cohesion is the undrained cohesion (cu)
i.e., a total stress approach.

It must be stressed that although this is the traditional approach to predicting pile
capacity, there would appear to be little theoretical justification. After all, the
majority of load borne by piles is often long-term load, and in any case, it is the
effective stresses in a soil which dictate its strength. The undrained approach has
probably been adopted because undrained strength tests of clay are quicker and
cheaper, and static load tests are in any case relatively quick, and therefore
represent an undrained condition.

3.1.1 The generalized pile capacity equation

The assumption of undrained conditions lead to a simplification of the general

equation to yield :

𝑃 𝑄 𝑄 𝐴 𝑓 𝐴 𝑓 𝐴 𝑐 𝑁 𝐴 𝛼𝑐
The values of Ab and As are determined by the pile geometry, however, the other
variables require careful examination. fb and fs are unit base and shaft
resistances of pile in kPa.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.14
Topic 3A: Axial Capacity of Piles

3.1.2 The Nc Parameter

10
The bearing capacity theories do not

Bending capacity factor Nc

suggest the adoption of different bearing
capacity factors, Nc, for displacement 9
(e.g. driven) and non-displacement (e.g.
drilled) piles. The factor will therefore be 8
treated here simultaneously for both
types of pile. 7

According to Terzaghi, the Nc 6

parameter for use in his bearing capacity
equation is a function of  only, and for 5
circular or square shallow foundations 0 1 2 3 4 5
(D/B < 1) has a value of about 6. L/d B
However, as was demonstrated by
Skempton (1951), Nc is also influenced by depth as is shown in this figure.

On the basis of Skempton's work, it is now common practice to use the value of
9 for all deep foundations in clay. As such foundations usually have a D/B ratio
much grater than 4, this value is nearly always appropriate. However, it is
important to note that where the D/B ratio is less than 4 (Bell & Robinson (2012)
reported this limit as 3), Nc should be reduced appropriately.

3.1.3 Nc in layered soils

Care should be taken when piles penetrate Ground surface

into stiffer clays underlying soft clays as
illustrated in this schematic. Although the
physical value of D/B may be very large, its
Soft
value relevant to the situation illustrated clay
D
should be taken as D'/B. This value could be
significantly less than 4 so that the Nc value
used reflects the fact that not all of the base
failure pattern is within a stiff clay.
Stiff B
Meyerhof (1976) defines a critical D
clay
penetration ratio, (D/B)crit which is the
penetration of a layer necessary to assume
point failure is entirely contained within that
layer, and not influenced by the softer upper layer. This critical ratio is defined
as 3 for  = 0, but increases to as much as 7 for  = 30o (which would be relevant
for mixed soils with both c and ).

The point bearing capacity in the lower layer could be assumed to increase
linearly from the full bearing capacity of the softer overlying soil at the surface of
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.15
Topic 3A: Axial Capacity of Piles

the underlying stiffer soil to the full bearing capacity of the stiffer soil at the critical
penetration ratio.

3.1.4 The value of cb

It was noted previously that there were many factors which contributed to the
differences between the cohesion measured during site investigations and the
value which is applicable to the base resistance of a pile in clay (see Difficulties
in application). Nevertheless, in the absence of any better value, the value of
cohesion determined from high quality laboratory or in-situ tests is used for driven
pile design.

The effects of pile installation, disturbance, porewater pressure development and

dissipation and consolidation are significant for soft clays. Nevertheless, soft
clays are obviously not used for end bearing support of piles. Piles may however
be founded in stiff or very stiff clays, which tend to be less sensitive to disturbance
effects. In these clays, excess porewater pressures are less dominant, and
variations of strength with time are minimal. It follows then that these influences
can be largely ignored.

The effect of fabric, and in particular the presence of fissures, may be ignored for
driven piles, but is important for bored piles. For driven piles, the driving process
causes substantial disturbance a remoulding of soils, which tends to destroy the
original structure of the clay.

However, for bored piles, where soil is removed, the remoulding of soil around
the pile does not occur. In the best-case scenario, the soil fabric remains intact,
and in the worst-case scenario (equivalent to poor construction techniques), the
fabric opens due to stress relief resulting from excavation. As noted previously
(see Scale effects), it is usual to incorporate a reduction factor to account for the
effect of fissures on the end bearing of bored piles, thus:

Qb  Ab c b N c

Where  = a reduction factor which relates the ratio of the full scale strength to
the smaller sample strength.

On the basis of comparative strengths determined by a 5.5mm diameter

penetration device, 38mm and 98mm triaxial tests and 865mm diameter plate
loading tests by Rowe (1972), it is suggested that triaxial test results should be
reduced to between 60% and 90% of their nominal value in fissured clays. By
comparison, Whittaker and Cooke (1966) recommend a reduction factor of 75%
for large bored piles in (fissured) London clay.

In general, it will not be possible to determine an appropriate reduction factor for

bored piles in fissured clays, so these recommendations should be used as a

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.16
Topic 3A: Axial Capacity of Piles

guide. An alternative conservative approach could be to adopt a lower-bound

strength profile from triaxial test results.

3.1.5 The value of cs

The value of shaft cohesion (cs), which is adopted in driven pile design is the
undrained value measured in conventional small scale laboratory or in-situ tests.
Any variation between this value and the value which is applicable to a full scale
driven pile is accounted for by the adhesion factor (). This adhesion factor is
determined empirically from the results of both full-scale and model pile load
tests, and ideally accounts for installation, disturbance, scale and loading rate
effects. There have been many such empirical correlations proposed but  is
usually given as a function of undrained cohesion (cu) and reduces with
increasing cohesion (see Poulos and Davis, 1980).

The adhesion factor () is a purely empirical factor which relates the likely pile
shaft resistance to values of undrained cohesion (cu) of undisturbed clay. There
is no theoretical basis for .

As discussed previously, there are many factors which can cause the cohesion
at the interface of a pile shaft to differ from the cohesion measured during site
investigations. However, it is generally assumed that the adhesion factor ()
takes all these factors into account. The problem therefore is to select an
adhesion factor which is appropriate for the particular soil and method of pile
installation. The following discussion will consider driven and bored piles
separately.

[Link] Adhesion factor for driven piles

As there is no theoretical basis to the value of , it is clear that it must be selected

from empirically based data derived from the results of pile tests conducted by
others. This,
however, Undrained shear strength (cu ) kN/m2
does not 50 100 150 200 250
provide a 2.0
very clear Figures denote penetration ratio =
Adhesion factor

picture as the Depth of penetration in clay

1.5 Pile diameter Key:
scatter of 49 49 Steel tube piles
results 19 49 56 Precast concrete
obtained 1.0 13 15 piles
13 17 27
44 39 Design curve for
from many 44 39 27
33
penetration ratio > 20
33
such tests is 0.5 19
38 40
17 10 5 35 44
very large. 15 8
This figure 0
shows a 1000 2000 3000 4000 5000
Undrained shear strength (cu ) lb/ft2
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.17
Topic 3A: Axial Capacity of Piles

number of results obtained for driven piles in clay.

It demonstrates that the adhesion factor can vary a great deal for a particular type
of pile installed under fairly similar conditions. A very courageous person has
indicated a design curve on this figure, but adopting this recommendation would
not necessarily guarantee a safe design. If a conservative approach were
adopted, a lower bound curve to the data set would result in grossly expensive
foundations. It would seem that some form of testing of the installed piles is
necessary to confirm capacity.

For soft insensitive clays, it is normal practice to adopt adhesion factors of about
unity, or even higher. This is because the immediate loss of strength (from
undisturbed to remoulded values) due to driving of the pile is compensated, or
even over compensated by the subsequent consolidation of the soil. However,
for sensitive soils, the immediate loss of strength will be much higher, and may
not be fully compensated by the subsequent consolidation. In such cases, an
adhesion factor of less than unity would be appropriate.

Based on the design experience in off-shore conditions, the adhesion factor can
be estimated in terms of overconsolidation ratio of the clay as a function of (cu/’v).

The variations of adhesion factor is shown in figure (a) as shown below (Semple
& Rigden, 1984; cited from Tomlinson & Woodward, 2008). For the most of the
onshore pile design, however, the effect of pile slenderness effect is minimal. For
L/B (i.e., length to diameter ratio) ratio exceeding 50, the earlier adhesion factor
is to be further reduced by multiplying it with a length factor (F) as per the figure
(b) below. The length factor will allow both for the flutter effects and the residual
or part-residual shear strength conditions at the interface.

American Petroleum Institute (API) in 2007 modified the adhesion factors of

Randolph & Murphy (1985) and provided the following equations for the average
shaft resistance:

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.18
Topic 3A: Axial Capacity of Piles

. . .
𝑓 0.5 𝑐 𝜎′ 𝑂𝑅 0.5 𝑐 𝜎′
[take the value whichever is larger]

Norwegian Geotechnical Institute (NGI) produced an alternative method for the

adhesion factor () known as NGI-99 method (Karlsrud et al. 2005). The method
takes into account the plasticity index (PI) of the clay in addition to OCR (refer to
Das & SIvakugan, 2019). The method is graphically presented in the figure below.
Based on the review by Doherty & Gavin (2011), the NGI method produced
comparable results with API for PI>20%.

[Link] Adhesion factor for bored piles

There are many influences on the adhesion factor of bored piles in clay that are
inherent in the method of construction. These are as follows:

 stress relief that occurs on the walls of the borehole to produce swelling and
weakening;
 migration of groundwater to this area;
 opening of fissures;
 formation of a slurry against the borehole walls due to water seepage;
 formation of a cake of bentonite if this is used in the drilling process;
 water migration from the in-situ concrete;
 shrinkage of the concrete.

A number of detailed investigations have been directed to a better understanding

of some of these effects, but it would be fair to comment that little general
progress has been made. Overall, however, it would appear for a wide range of
conditions for tests conducted on bored and cast-in-situ piles at a number of

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.19
Topic 3A: Axial Capacity of Piles

locations about the world, the adhesion factors generally range from 0.3 to 0.6
(Tomlinson & Woodward, 2007).

Perhaps the best known of these is the study by Skempton (1959), in which it
was found that the average adhesion for London clay was about 0.45 for normal
length piles, which were drilled and filled with concrete soon after. For short piles,
the adhesion factor was about 0.3 is suggested- the reason given being that
heavily fissured clay reduced the resistance (perhaps in pullout the failure was
through the fissures rather than at the pile-soil interface). Skempton also found
that the shaft resistance did not exceed about 100 kPa. For a small diameter
bored piles and piles with enlarged base where long delay beween drilling and
concreting can take place, an adhesion afctor of 0.3 is to be used. If the enlarged
base is to be constructd in a fissured clay, Tominson & Wooodward (2007)
suggested ignoring 2B length of shaft friction behind the enlarge base.

Unfortunately, there appears to be little guidance on what adhesion factor should

be taken. While for London clay,  = 0.45 appears to be acceptable, a more
reasonable approach may be to use a lower value of about 0.3 initially, and to
confirm this later or increase it as a result of load testing. The shaft resistance
should not exceed 100 kPa.

3.2 Effective stress approach (ESA) for shaft resistance in clay

As mentioned previously, the total stress approach (using undrained strengths)

for pile capacity determination has little theoretical justification. It has been shown
that there are a number of installation effects (e.g. remoulding, stress increase,
stress relief, fissuring, carry down of layers, consolidation) as well as soil strength
and sensitivity which can affect the pile capacity. For shaft resistance, all these
effects are unrealistically combined in the empirical adhesion factor, .

On the other hand, an approach based on effective strength parameters would

seem to have a good theoretical basis. If the effective strength parameters can
be established, the base resistance can be obtained using the appropriate values
in the bearing equation, and the shaft resistance (e.g., -method) will be a
function of the horizontal stress at the pile-soil interface. Unfortunately, this
approach introduces its own set of complicated and poorly understood factors.

3.2.1 The generalized pile capacity equation in clay based on ESA

The effective strength parameters for cohesive materials suggest that the
strength of the clay is stress-dependent i.e., the more stress applied, the denser,
and therefore stronger will be the clay in the long term. The (stress-independent)
cohesion value we used for the total stress approach can now be assumed to be
zero, and the general pile equation can be simplified to:

𝑃 𝑄 𝑄 𝐴 𝑓 𝐴 𝑓 𝐴 𝑃′ 𝑁 𝐴 𝐾𝑃′ 𝑡𝑎𝑛𝛿

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.20
Topic 3A: Axial Capacity of Piles

It should be noted that the assumption that c' = 0 may be reasonable for soft clays
but for stiff clays, a small but important cohesion is generally measured. The
assumption may therefore be an oversimplification.

[Link] Shaft resistance in clay based on ESA

With regard to the side resistance component, the unit shaft resitance (fs) can be
determined based on the effective stress parameters as:

𝑓 𝐾𝑃′ 𝑡𝑎𝑛𝛿 𝛽𝑃′ Where, =Ktan Ktan' (If it is assumed that

=’).
Burland (1973) assumed that the K value could be reasonably represented by
Ko, the "at rest" earth pressure coefficient.

For normally consolidated (NC) clay: 𝐾 𝐾 1 𝑠𝑖𝑛𝜙

For overconsolidated (OC) clay: 𝐾 𝐾 1 𝑠𝑖𝑛𝜙 √𝑂𝐶𝑅

It may be of interest to note that the above relationships give an effective stress
adhesion factor (), which is defined by

This should be compared with the total stress adhesion factor (), which is
defined by 𝛼 .

For most types of clay, ' varies between approximately 20 and 40, although it
must be stated that values much less than 20 are fairly common especially for
the more active clays. Mitchell (1976) has proposed the following relationship for
the critical void ratio (i.e. residual) friction angle for clays:

sin cv
  0.8  0.094 lnIP

Where Ip=plasticity index

There may be some justification for using a residual friction angle in pile design
given the disturbance caused by installation. If the clay is sensitive, and
remoulding of the soil has not occurred (e.g. in bored piles with careful
construction control), higher values can perhaps be considered.

A range of 6 < IP < 100, yields values of ' between 39 and 21.5,

Substitution of these values yields values of  in the range

0.25    0.30
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.21
Topic 3A: Axial Capacity of Piles

The above range is relatively narrow, indicating that  is not highly dependent on
clay type.

Parry and Swain (1977) adopted a similar approach and derived a slightly more
complex equation for , namely

sin ' cos '


1  sin 2'

For 21.5  '  39, yields  values of

0.30    0.35

This range is a little higher than derived from Burland's equation and does seem
to fit the test data a little better.

However, while there does seem to have been some form of success with soft,
normally consolidated clays, the same cannot be said for stiff, overconsolidated
soils. The reason for this is that the K value can be very variable and its
relationship to Ko is greatly influenced by the method of installation. The value of
Ko can also vary a great deal in overconsolidated clays.

By assuming K = Ko for driven piles, for the known distribution of Ko in London

clay, Burland demonstrated that the effective stress approach provided a lower
bound to the test results obtained. The driven piles presumably demonstrated
greater shaft friction than predicted because K > Ko as a result of lateral
displacement during pile installation.

By contrast, for bored piles in stiff London clay, it was found that the effective
stress method provides an upper bound to the shaft friction measured in field pile
load tests. The bored piles presumably demonstrated lower shaft friction than
predicted because K < Ko as a result of stress relief during drilling.

These general trends are supported by Meyerhof (1976), who demonstrated that
for driven piles in stiff clay, K  1.5Ko, while for bored piles, K  0.75Ko.

3.2.2 Base resistance of clay based on ESA

For the base resistance component, there appears to be very little available in
the literature. However, according to Poulos and Davis (1980), it would appear
reasonable to use the effective overburden pressure at the toe of the pile, P'ob
along with a value of Nq that has been estimated from a knowledge of the
effective angle of friction in the same manner as granular soils. However, while
in principle this approach seems reasonable, it should be handled with some care
as it has not really been proven in engineering practice.

Department of Civil Engineering, Monash University

Revised: 6/2020
 CIV4249: Foundation Engineering 3A.22
Topic 3A: Axial Capacity of Piles

3.2.3 Closing remarks

In summary, while the effective stress approach appears to be theoretically sound

for piles in clay, it can be difficult to apply with confidence, as the normal stress
at the pile-soil interface is highly dependent on the pre-existing stress conditions,
reflected in the parameter Ko as well as the post-installation stress state reflected
in the parameter K. This parameter is particularly difficult to establish in
overconsolidated clays.

3.3 The lambda method

An alternative and entirely empirical method has been proposed by Vijayvergiya

and Focht (1972) for the estimation of the side resistance of long steel pipe piles
founded in clay based on total stress (undrained) method. This method is used
fairly frequently in the design of heavily loaded offshore foundations. Because
these piles are long and slender, the great majority of capacity is derived from the
shaft and, therefore, the end bearing component can be insignificant.

This method is not commonly used for land-based piles, and should only be
applied where an assumption of normal consolidation is appropriate.

 The authors simply

0 0.1 0.2 0.3 0.4 0.5 established a correlation
between ultimate shaft
resistance (Qs) determined
25 from a large number of load
tests on steep pipe piles, the
50 mean effective vertical stress
between ground and pile toe
Psu
= ( m + 2cm )As
('m) and the mean
75
undrained cohesion along the
Pile penetration ( ft. )

Location Symbol Source pile shaft (cm) as follows :

100
Detroit Housel
Morganza Mansur
Cleveland Peck Qs  (m
  2cm )As
125 Drayton Peck
North Sea Fox
Lemoore Woodward Where,  is a
150 Stanmore Tomlinson dimensionless coefficient.
New Orleans Blessey
Venice McClelland
Alliance McClelland
From all the pile test results
175 Donaldsonville considered, a reasonably well
Darragh
MSC Houston McClelland defined variation in the
San Francisco Seed
200 British Columbia McCammon parameter  was revealed
Burnside Peck and is reproduced in this
figure.
225

It follows then that  is a

function of pile penetration and decreases to a reasonably constant value for very
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.23
Topic 3A: Axial Capacity of Piles

large penetrations. It is possible to compare the conventional adhesion factor ()

with  from a comparison of the relevant equations. Thus

As c m  (m
  2c m )As

From which it follows:

  
    m  2
 cm 

It must be borne in mind that the lambda method was proposed for marine pile
applications, for which an assumption of normal consolidation is usually
appropriate. For normally consolidated clays, it is not unreasonable to find that
the ratio of undrained cohesion to effective vertical stress is of the order of about
one third (Skempton, 1957), i.e.

cm 1

v 3

By substitution we can derive

  5
The accuracy of this can be established in relation to the typical
recommendations for shown previously. A stiff clay has an undrained cohesion
of about 100 kPa. It follows from the above and a consideration of likely unit
weights of a normally consolidated clay, that the stiff consistency should be
encountered at a depth of about 60 m (or about 200 ft). From the above figure, 
is about 0.1. From the foregoing, it would appear that this is equivalent to an
adhesion factor ( of about 0.5, which is indeed quite reasonable for a stiff clay.

In a similar manner, taking the undrained cohesion of a soft clay as about 25 kPa,
such a strength should be encountered at about 15 m (50 ft.). At this depth,  is
about 0.2 and this corresponds to an  value of about 1. Once again, the two
techniques seem to agree fairly well.

4. Single piles in granular soils

For granular soils, drainage is generally so fast that the distinction between
drained and undrained or effective and total stress analysis never arises. As with
piles in clays, the general pile capacity equation may be relatively simple but the
specific values used in the equation depend on a great number of factors and
are, as usual, difficult to evaluate.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.24
Topic 3A: Axial Capacity of Piles

Driven piles are a common solution in granular soils. The driving process
densifies and strengthens a granular soil with the result that good pile
performance can be expected.

Bored piles, on the other hand, cannot be constructed in granular soils without
some form of temporary or permanent support. Without such support, the sides
can collapse and the base may heave. Support can be provided by temporary or
permanent casing, which can be driven or screwed into the ground prior to drilling,
or by use of drilling muds (e.g. bentonite/polymer fluid) which apply a stabilizing
hydrostatic stress to the walls. The former solution can be expensive, the second
solution is cheaper, but requires considerable expertise, and is messy. It should
also be borne in mind that the remnant bentonite/polymer may reduce the
available shaft resistance for the pile.

4.1 The generalized pile capacity equation

For granular soils, it can be assumed that the cohesion is zero and therefore, the
generalized pile capacity equation reduces to (similar to the ESA for clay):

𝑃 𝑄 𝑄 𝐴 𝑓 𝐴 𝑓 𝐴 𝑃′ 𝑁 𝐴 𝐾𝑃′ 𝑡𝑎𝑛𝛿

4.1.1 The Nq parameter

For shallow foundations (D/B < 1), it is normally assumed that Nq is a function of
’ only (Terzaghi and Peck, 1967). However, as D/B increases above 1, it is
generally accepted that the Terzaghi and Peck bearing capacity factors become
increasingly conservative (c.f. the effect of D/B on Nc for piles in clay). There
have been a number of suggestions for the Nq proposed over the years. Some
of these, and the methods for their implementation, will be discussed hereafter.

4.1.2 Meyerhof’s method

Meyerhof (1951, 1976) proposed bearing capacity factors, which can be found in
most texts [Bowles (1997), Das & Sivakugan (2019)]. These bearing capacity
factors depend on the pile having reached a penetration ratio (D/B) greater than
a critical value, (D/Bcrit) which is a function of the friction angle . Graphs of the
critical penetration ratio will be found together with Meyerhof’s Nq values. Typical
values for a range of friction angles are:

 = 35o, D/Bcrit = 11 and Nq = 143;

 = 40o, D/Bcrit = 17 and Nq = 346;

 = 45o, D/Bcrit = 25 and Nq = 930.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.25
Topic 3A: Axial Capacity of Piles

It is important to note that these values of  are the in-situ undisturbed values.
Some of the older methods of design have suggested that significant reductions
should be made to the measured ' value when designing bored piles to account
for loosening effects. Although Meyerhof’s method does not explicitly allow for
differences between driven and bored piles, consideration should be given to
reducing the value of ' to between 0.5 and 0.7 if loosening of the soil is
considered likely.

If the critical embedment ratio is not attained, these values must be reduced
accordingly.

For a homogenous soil, Meyerhof would suggest a linear increase in bearing

resistance with depth from zero at the ground surface to the maximum value at
the critical embedment ratio. An alternative approach would be to analyze the pile
in this case as a shallow footing, and use the appropriate bearing capacity factors.

For layered soils, where the pile penetrates from a loose soil into a dense soil,
Meyerhof suggests that for all practical purposes, the full bearing capacity of the
denser soil can be assumed if the penetration exceeds 10 pile widths (D/B > 10).
Presuming that the total pile embedment still satisfies the critical ratios mentioned
before, Meyerhof recommends the following reduced unit point bearing pressure:

q b  q l (l ) 
ql (d )  ql (l ) Dd  ql (d )
10B
Where, qb is the unit base resistance;
ql(l) is the limiting unit base resistance in the loose sand based on the
appropriate values of Nq and  for the loose sand;
ql(d) is the limiting unit base resistance in the dense sand based on the
appropriate values of Nq and  for the dense sand;
Dd is the pile penetration into the dense sand, and
B is the pile width.

N.B. Meyerhof’s method is used with the (unlimited) effective overburden

pressure, but is also subject to an overriding limiting point resistance. (See under
The value of P'ob).

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.26
Topic 3A: Axial Capacity of Piles

4.1.3 Berezantzev’s method

Berezantev et al. (1961) method is widely used to determine Nq for piles in

granular soils. His proposed values are relatively insensitive to the effect of
embedment, and they are therefore generally assumed to be independent of
embedment ratio (D/B).
However, the Nq values
are highly sensitive to
friction angles. The
variation of Nq with ’ is
shown here (Bell &
Robinson, 2012).

A major difference in the

application of
Berezantev’s method is
that the value of friction
angle (') used is the
post-installation value.
The value of '
determined from various
in-situ tests such as the
SPT or CPT are
appropriate to the soil condition before the installation of piles, and do not include
the influence of pile construction.

A number of methods have been proposed, but one of the simplest, and indeed
one that is accepted generally, has been proposed by Kishida (1967). This
method suggests that the value of ’ to be used for the determination of Nq for
driven piles is:

 '  0 .5 (  ' 1  4 0 )

Where ’1 is the representative value of ' at the pile toe before pile installation.

Similarly, Kishida proposes that for bored piles, the value of ' to be used for the
determination of Nq is

   1  3

Some typical bearing capacity factors determined by this method are:

 = 35o, Nq driven = 120 Nq drilled = 45;

 = 40o, Nq driven = 200 Nq drilled = 110;

 = 45o, Nq driven = 300 Nq drilled = 250.

Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.27
Topic 3A: Axial Capacity of Piles

N.B. Berezantev’s method is used in combination with limiting depth theory, and
effective overburden pressures reduced from their theoretical values. (See under
The value of P'ob).

4.2 The value of P'ob

The value of effective overburden pressure at the pile base would appear to be a
simple enough calculation, however, there is great debate in the geotechnical
community about the appropriate value. This section will describe some of the
alternative postulations.

In many of the older methods of pile design, it has often been assumed that the
vertical effective stress that is used in pile design in a granular soil increases
linearly with depth just as the vertical effective overburden pressure (Meyerhof,
1951 and 1976).

Later, it was postulated by researchers such as Vesic (1967) on the basis of

model tests, that the vertical effective stress reaches a limiting value at a certain
(critical or limiting) depth, beyond which there is assumed to be no increase (see
figure). This seems to be borne out by full-scale pile load tests. The limiting
vertical (and therefore horizontal) stress effect has been attributed to arching in
the soil and particle crushing.

W.T.
zc

vc

Most recently, some researchers (Kulhawy, 1984) have returned to the idea of a
vertical effective stress that is not limited by some arbitrary critical depth. Others,

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.28
Topic 3A: Axial Capacity of Piles

(Randolph, 1993) have postulated that a bulb of high stress follows the pile tip
down during driving, and that the shaft resistance behind the tip is progressively
fatigued. The jury is still out! This highlights that in geomechanics we are dealing
with the complex interaction of man-made objects and nature - this is ultimately
what makes geomechanics continually interesting and challenging!

4.2.1 Unlimited depth method - Meyerhof

Meyerhof (1951) proposed an alternative set of bearing capacity factors, N'c and
N'q for deep foundations. These are given graphically in Bowles (1988) or Das &
Sivakugan (2019). The method of Meyerhof allows full calculation of effective
overburden pressure at the base, in combination with the Meyerhof bearing
capacity factors for deep foundations (see Meyerhof’s method).

Note, however, that Meyerhof also proposed limiting values of end bearing which
could not be exceeded by the product of Nq and P'ob, i.e.

Qb  Ab 50N q tan 

This translates to maximum unit end bearing values as follows:

 = 30 (loose), qb  1.7 MPa;

 = 35 (medium), qb  4.5 MPa;
 = 40 (dense), qb  14 MPa;
 = 45 (very dense), qb  45 MPa;

N.B. In general practice, it is uncommon to allow ultimate end bearing

values in excess of 15 MPa. 20
4.2.2 Limited depth method – Vesic
(1967)
15
Based on the test results obtained by
Vesic (1967), Poulos and Davis (1980)
zc / d

have developed a means of estimating

the critical depth, zc , in the form of the 10
ratio zc/B, as a function of the effective
angle of friction, '. This is shown in the
accompanying figure. It should be noted 5
that the value of ' used in this figure
should be the value that has been
adjusted to reflect the method of 0
installation as follows: 28 33 38 43

Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.29
Topic 3A: Axial Capacity of Piles

For driven piles:   0.751  10

For bored piles:   1  3

Once the value of zc/B is known, the value of P'ob can be calculated. If the toe
of the pile is above this critical depth, then the value of P'ob is simply the vertical
effective overburden pressure. However, if the toe of the pile below this critical
depth, P'ob must be limited to the effective vertical overburden pressure at the
critical depth.

  z
Pob for z  z c
i.e.
P 'ob = z c  for z  z c

4.3 Theory of expansion of cavity-Vesic’s method

Vesic (1977) proposed a method for estimating the pile end bearing capacity
based on the theory of expansion of cavity (refer Das & Sivakugan, 2019 for
details). As per this method, the end bearing resistance in sand (fb) is calculated
as:

𝑓 𝜎′ 𝑁 ∗

Where, 𝜎′ =mean effective normal stress at the pile toe level= 𝑞′

q’=effective stress at pile tip level; Ko=earth pressure at-rest=1-sin’

𝑁 ∗ =bearing capacity factor=f(Irr)

Irr=reduced rigidity index= where, =average volumetric strain in the plastic

zone below pile tip

Ir=rigidity index= where, Es= Elastic modulus of soil,

Gs=shear modulus

𝜇 0.1 0.3 for 25o≤’≤45o

Δ 0.005 1 where pa=atmospheric pressure (100 kPa)

Refer to Das & Sivakugan (2019) for 𝑁 ∗ values with friction angle (’) and
reduced rigidity index (Irr).

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.30
Topic 3A: Axial Capacity of Piles

4.4 Kulhawy’s Method

As noted previously, Kulhawy (1984) suggested that the limiting depth approach
to determination of pile capacity in sand was a fallacy which resulted from the
effects of overconsolidation on the capacity of piles of typical (<20m) length. What
is more, the limiting depth theory was based on the results of model pile tests
alone. He proposed that the full overburden pressure should be used, and the
unit tip resistance be computed as:

q b  0.3BN   r  qN q qr qs qd

For D/B > 4~5, the first term becomes less than 10% of the second term, and can
be ignored. The dimensionless portions of both terms are given by Kulhawy in
graphical form against friction angle, . To use these graphs, it is necessary to
estimate the so-called rigidity index, Ir, which he approximates as follows:

30
 in loose sand, I r 
2
where q is in ton / ft
1/ 2 o o
q tan(28 to 26 )
110
 in dense sand, I r 
2
where q is in ton / ft
1/ 2 o o
q tan(40 to 37 )

The rigidity index can also be computed from first principles.

4.5 The value of P'os

As for the determination of effective overburden at the pile base P'ob, there is
dispute as to whether the full or limited overburden pressure should be used for
computing the effective overburden pressure for shaft resistance, P'os. The
following summarizes the different methods

 Meyerhof (1976) - use full calculated effective overburden pressure

 Kulhawy (1984) - use full calculated effective overburden pressure

 Vesic (1967) use full calculated effective overburden pressure above critical
depth, and limited effective overburden pressure below the critical depth. (See
also previous).

 In this method, the following post-installation friction angle, ', should be used
to compute the critical depth ratio, zc / d:

For driven piles : ' = 0.75 '1+10

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.31
Topic 3A: Axial Capacity of Piles

For bored piles :    1  3

where '1 is the undisturbed friction angle.

 The end-bearing capacity of a pile in a uniform layer of coarse-grained soils

will not continue to increase proportionately with depth. Vesic (1977) reported
that the end bearing resistance reaches to a limiting value typically for pile
lengths in excess of about 20D (where D is the pile diameter). For this reason,
the end-bearing is often limited conservatively to between 10 and 15 MPa for
dense sands or gravels, and to about half these values for medium-dense
sands or gravels (Bell & Robinson, 2012). Pile installation method also needs
to be accounted for.

4.6 The parameters K and 

The parameters K and  are sometimes treated separately, but in other design
methods are combined for convenience. This section will initially look at these
parameters individually, and then discuss the methods which use the combined
parameters K tan  or K tan .

4.6.1 The lateral pressure coefficient K

The value of K must be related to the initial horizontal earth pressure coefficient
(Ko = 1 - sin ), however, because of the very variable effects of pile installation,
the relationship is far from clear.

Tomlinson & Woodward (2008) recommends the following typical values for K /
Ko:

Pile Type K / Ko

Small-displacement, driven piles 0.75 to 1.25

Large-displacement, driven piles 1 to 2
Bored and cast-in-palce piles 0.7 to 1
Jetted piles 0.5 to 0.7

4.6.2 The pile-soil friction angle, 

The angle of friction between the pile shaft and the soil () has an upper limiting
value of '. Clearly, this upper limit would apply when the side of the pile is so
rough that the shearing occurs entirely within the soil. However, for pile shafts
which are smoother, shearing will occur at the interface with a friction angle  < 
'. It follows that  is a function of the surface characteristics of the pile shaft. There
have been many suggestion as to the values of  which should be used for
different surface finishes.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.32
Topic 3A: Axial Capacity of Piles

Kulhawy (1984) recommends the following typical values for /’:

Pile/soil interface / '

Smooth (coated) steel/sand 0.5 to 0.7

Rough (corrugated) steel/sand 0.7 to 0.9
Precast concrete/sand 0.8 to 1.0
Cast-in-place concrete/sand 1.0
Timber/sand 0.8 to 0.9
4.6.3 The combined parameter K tan 

Some design methods (Meyerhof and Vesic among them) consider the two
factors together as the function (K tan  but sometimes denoted K tan ). These
methods will be outlined in the following:

[Link] Meyerhof’s method

Meyerhof (1976) has proposed values 1.6

of K tan  for driven, jacked and bored
piles. These are shown graphically in
Poulos and Davis (1980) and is 1.2
reproduced here. Note that the shaft Driven
K s tan a

resistance values reflect the likely piles

changes of stress state in the soil due 0.8
to the method of installation. The
values for bored piles are based on an Jacked
piles
assumed reduced friction angle ' of
75% of its undisturbed value. In using 0.4
this chart, the undisturbed value is Bored
used in all cases. piles
0
These values are combined in the 28 33 38 43
Meyerhof method with the full 1
calculated effective overburden
pressure.

Meyerhof also proposed some simple guidelines for quick estimation of shaft
resistance for driven piles based on SPT N values :

 high displacement driven piles qs  2N (in kPa)

 low displacement driven piles qs  N (in kPa)

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.33
Topic 3A: Axial Capacity of Piles

[Link] Vesic’s method

On the basis of Vesic's (1967) test results, 3.0

Poulos and Davis (1980) have suggested
a relationship between K tan  and ' as
shown here. The value of ' is the post- 2.5
driving value (see previous). It should be
noted that Vesic's tests were conducted

K s tan a
on steel piles, and it is likely that the
values given in this figure are 2.0
conservative for other (rougher) surface
finishes.
1.5
The values of K tan  given for Vesic’s
method should be compared with those
derived from Meyerhof's (1976) results 1.0
shown previously. It will be seen that the 28 33 38 43
K tan  values for Vesic’s method are 
about twice those of Meyerhof. Note,
however, that different overburden pressures apply. Poulos and Davis suggest
that Vesic’s method is preferred for driven piles, but recommend using Meyerhof’s
method for bored piles.

Bell & Robinson (2012) summaried the -values of different researchers for
driven and bored piles as given below:

Material -value (Driven pile) Reference

Silica sand β = 0.15–0.35 (compression) McClelland (1974)

β = 0.10–0.24 (tension)
β = 0.44 for ′ = 28° Meyerhof (1976)
β = 0.75 for ′ = 35°
β = 1.2 for ′ = 37o

Uncemented  = 0.05–0.1 Poulos (1988)

calcareous sand

Material -value (Bored pile) Reference

Silica sand  = 0.1 for ′ = 33° Meyerhof (1976)

 = 0.2 for ′ = 35°
 = 0.35 for ′ = 37°
Kraft and Lyons
 = F tan (′- 5)
(1974)
where F = 0.7 (compression) &
F = 0.5 (tension)

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.34
Topic 3A: Axial Capacity of Piles

Uncemented  = 0.5–0.8 Poulos (1988)

calcareous sand fslim=limiting shaft resistance =
60 to 100 kPa

N.B. For long piles in coarse-grained soils, average shaft friction is to be limited
110 kPa unless otherwise proven by suitable pile load testing or experience in
similar ground conditions.

[Link] Stabilization of bored holes

It is usually necessary to employ some form of side support system when

installing bored piles in granular soils. When casing is used, it is normal to apply
the above design recommendations.

A more common (i.e. cost-effective) technique is to use bentonite slurry/polymer

to support the drilled hole during excavation. The bentonite is displaced by
pouring the concrete through a tremie pipe from the base of the pile up.
Theoretically, any slurry caked on the sides is meant to be scoured off by the
action of the rising concrete, however, this may not always occur in practice.
Sliwinski and Fleming (1974) suggest that the ultimate shaft resistance should be
reduced by about 10 to 30% to account for the bentonite, however, Touma and
Reese, (1974) suggest no reduction in shaft resistance.

[Link] Calcareous sands

The previous design methods apply to granular soils comprising silica sands -
these are the predominant on-shore sands. In the offshore environment,
calcareous sands predominate, and these behave entirely differently. Although
the angle of friction of calcareous sands prior to pile installation may be in excess
of 35 or even 45, the driving process tends to crush the sand particles
significantly. As a result, normal stresses at the interface are extremely low, with
the result that shaft resistances can be as low as 5 kPa.

This is particularly relevant to Australian offshore conditions where there are

significant deposits of calcareous sands in both the uncemented and weakly
cemented states (e.g. Bass Strait, Timor Sea and the North-West Shelf). Under
these circumstances, there may be considerable advantages in using pile
installation techniques other than purely driven piles. Pile types such as bored-
and-grouted, and driven-and-grouted piles have been used successfully. A
detailed discussion on these and other problems associated with calcareous
sands may be found in many papers contained in Jewell and Khorshid (1988).

4.7 Direct correlations with insitu SPT/CPT data

There are many forms of in-situ tests, which can be used as part of the overall
design process for any particular type of foundation. Many of these tests are used
to provide conventional soil parameters either directly (e.g. the pressuremeter) or
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.35
Topic 3A: Axial Capacity of Piles

indirectly (e.g. SPT/CPT) so that the design methods discussed above can be
used.

On the basis of a great number of empirical studies, however, some direct

methods of calculating pile shaft and end bearing have been developed. The CPT
is in fact like a mini pile, and the cone resistance and sleeve friction are analogous
to the end bearing and shaft resistance. The main consideration must be,
however, that the actual and model piles are significantly different in scale, and
the zone of influence of the actual pile must be taken into account.

4.7.1 Shaft and end bearing of piles based on SPT

Poulos (1989) proposed a direct relationship between the unit ultimate shaft
resistance of a pile with the SPT blow count as:

𝑓 𝛼 𝛽𝑁 𝑖𝑛 𝑘𝑃𝑎
Where, fs = unit ultimate shaft friction; N = SPT value, and α and β are constants
depending upon soil and pile type (refer to the table below).

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.36
Topic 3A: Axial Capacity of Piles

Poulos (1989) also proposed a direct relationship between SPT N value and unit
ultimate end bearing as:

𝑓 𝐾𝑁 𝑖𝑛 𝑀𝑃𝑎

Where, K=constant (refer to table below), values depend on pile type and soil.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.37
Topic 3A: Axial Capacity of Piles

For refernces used in the able tables refer to Poulos (1989).

Meyerhof (1976) proposed direct correlations between SPT N-value and unit
shaft/base resitances of a pile as:

Unit shaft resistance:

For driven piles: 𝑓 2𝑁 𝑖𝑛 𝑀𝑃𝑎

For bored piles: 𝑓 𝑁 𝑖𝑛 𝑀𝑃𝑎

Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.38
Topic 3A: Axial Capacity of Piles

Unit base resistance for pile penetration > 6D:

For driven piles in sand: 𝑓 0.4𝑁 𝑖𝑛 𝑀𝑃𝑎

For driven piles in silt: 𝑓 0.3𝑁 𝑖𝑛 𝑀𝑃𝑎

Where, N60=average N-value 3D above/below pile tip

4.7.2 Shaft and end bearing for piles based on CPT

The use of CPT for direct correlations with shaft/base resistances of driven piles
is discussed below. It is to be noted that the use of CPT data for bored piles is
highly unreliable unless the CPT results have been adjusted to account for the
changed ground condition affected by the pile installation methods. As such direct
correlations between CPT and bored piles are avoided here.

[Link] Shaft resistance - driven piles

It would appear that the base resistance of a driven pile in sand is directly related
to the cone resistance, qc.

i.e. Qb  Ab qc

For a uniform sand, this would be a relatively simple method of design. The main
problem arises when dealing with 2
natural sands, which can show Cone resistance, kg/cm
100
substantial variations in qc over fairly
short distances. It therefore becomes
largely a scaling problem so that the
representative value of qc can be
Readings at 20cm
extrapolated from the size of the intervals
Idealised curve of
cone resistance
Depth , m

penetrometer to the much larger size of A

the pile. There are a number of B
methods which have been suggested 60 cm
for this, all of which are fairly similar but
600 cm

varying in the details of the length over

which a representative qc is taken and
D
the way in which this is done. One of
the common methods for this C
procedure has been proposed by
Heijnen (1974) as shown in this figure.

There is another important difference between a CPT and a full-size pile which
concerns the depth which a pile must be driven into a bearing layer to achieve
the resistance predicted from the penetrometer results.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.39
Topic 3A: Axial Capacity of Piles

Suppose, as is shown above, the penetrometer tip first enters a bearing point
layer at point A. Because of the scale of the device, it may only require 350 mm
(10 diameters) of further penetration before the resistance of the tip fully reflects
the increased bearing strength.

In the case of a pile of 350mm, the pile penetration necessary to develop the
layer's full capacity may still be 10 diameters, but this would now be 3.5m as
shown in the figure. Therefore, if the required pile resistance was 100 kg/cm2 (9.8
MPa), then the pile must be driven to the depth of point D (which is approaching
a further 3 m of penetration) rather than to some point just.

This procedure was first proposed by De Beer (1963) but seems to give some
very large penetration which according to Vesic (1970) are uneconomical. Vesic
suggests that driving the pile a further 5 diameters after the depth at which the
peak cone tip resistance was recorded is sufficient. The difference is probably
due to the greater densification caused by the full sized pile.

White & Bolton (2005) suggested a direct correlation between ultimate unit base
(fbu) resistance and qc as:

𝑓 0.9𝑞
Where, 𝑞 =weighted average qc value 8D above and 4D below pile tip (Dutch
method).

[Link] Shaft resistance - driven piles

Once again, there are a great number of different methods for calculating the
shaft resistance of piles directly from CPT results; some based on sleeve friction
(fs) but most based on tip resistance (qc). High degree of caution is required
when using fs as the difference could be significant due to the soil consolidation
effect. An example of the former approach is provided by the following equation:

Qs = S1Asfs

where S1 depends on type of pile as shown in the table given below.

Type of Pile S1

Timber 1.2
Parallel-sided concrete or steel;
flat end 0.6
pointed end 1.1
Driven cast-in-situ* 1.6
Open steel tube and H-pile 0.7

* S1 may be higher if the concrete is rammed as the casing tube is withdrawn

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.40
Topic 3A: Axial Capacity of Piles

Begemann (1977) has suggested that the factor S1 should be taken as 0.7 for all
types of driven pile. It is also normal to have an upper limit for shaft resistance
which can vary, but according to Dutch practice (Meigh, 1987), is 0.12 MPa or
120 kPa.

As noted above, there are a number of methods for the calculation of shaft
resistance from cone tip resistance. Another method used is:

Qs = S 2 As qc

where S2 is given in the following table:

Type of Pile S2

Timber 0.012
Precast concrete 0.012
Precast concrete, enlarged base* 0.018
Steel displacement 0.012
Open-ended steel tube 0.008

* only applicable in a dense group of piles (otherwise reduce to 0.003

where shaft is narrower than base).

As before, a limiting of fs is often adopted and 0.12 MPa seems reasonable. It

may be worth noting that the values of S2 are reasonably consistent with the
corresponding values of S1 given previously, with for sands, the friction ratio
being typically 1%.

5. Underlying weak layers

It should be noted that if a weak layer

exists below the toe of a pile as is
illustrated here, the base resistance
could be adversely affected. This
would apply to both sands or clays. B
Meyerhof (1976) has suggested that if
the weak layer is within 10 pile widths,
zo/B < 10, then the base capacity Sand zo
should be reduced linearly from the
greater capacity at zo/B = 10 to the
Clay
weaker layer capacity at zo/B = 0.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.41
Topic 3A: Axial Capacity of Piles

6. Single piles in rock

The design of piles on and in rock is generally based on experience, empirical

rules or city ordinances. A method by Goodman (1980) can also be used.
Research at Monash also resulted in a fully theoretical solution (Seidel, 1993) for
the capacity of shafts drilled into rock. A recent paper by Haberfield & Lochaden
(2019) further highlights various design issues with the rock socket pile. It is to be
noted that the capacity of the rock socket piles primarily depends on the
serviceability limit state rather than the ultimate limit state requirement. However,
when designing a rock socket pile, both the ULS and SLS must be satisfied as
design requirements.

It is important to check that the calculated allowable load does not exceed the
structural capacity of the pile shaft.

6.1 End bearing

On the basis of a comparison of achieved end bearing and unconfined

compressive strength of intact rock samples (quc), the following end bearing
values are recommended. The strength of the rock material needs to be
compatible with the rock mass (e.g. joint, infilling, joint orientation, joint aperture,
solution features, weaker underlying strata).

𝑓 0.5𝑞 where, fb(all) is the allowable unit point resistance (FoS=2)

𝑓 𝑞 where, fbu is the ultimate unit point resistance

This recommendation is based not so much on strength criteria as on the

choosing an allowable strength compatible with structural deflection
requirements. Caution should be used when piles are in swelling shales.

If driven or bored piles are driven or placed in rocks where a practical refusal is
expected then the structural capacity of the pile governs the capacity. In the U.K.,
a maximum allowable limit of 0.25f’c for concrete piles is considered when
calculating the pile capacity. This ensures pile materials are not overstressed
from driving. If steel pile is used then the driving stress needs to be limited to
values, which will not cause damage to the rock below the tip.

6.1.1 Goodman’s method

Goodman recommends for piles driven onto rock that the following equation
applies:

𝑄 𝑞 𝑁 1 𝐴

Where Qp is the ultimate point resistance

quc is the mass unconfined compressive strength
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.42
Topic 3A: Axial Capacity of Piles

Nq = tan2 (45 +  / 2)
 is the drained angle of friction of the intact rock

Goodman recommends that any laboratory estimate of unconfined compressive

strength be divided by 5 in order to account for the fabric (discontinuities) in the
rock mass. In addition, a factor of safety of 3 is generally applied to determine the
allowable point bearing.

Typical values of unconfined compressive strength for different rocks can be

found in Poulos and Davis (1980), and in Das & Sivakugan (2019). Das &
Sivakugan also tabulates representative values of angles of internal friction for
many different rock types.

6.1.2 Zhang’s method

Zhang (2010) introduced a method for determining the end bearing resistance
based on the unconfined strength of rock mass as:
.
𝑓 6.39 𝜎
where, 𝜎 is the rock mass unconfined compressive strength =()0.7c, c=UCS
of intact rock, E=0.0231RQD-1.32 ≥ 0.15; E = reduction factor to relate intact
rock UCS with rock mass UCS.

Zhang also summarised in the paper a number of methods available to relate the
UCS of intact rock with the UCS of rock mass based on various rock mass
classification systems (i.e., RMR, GSI, Q).

6.2 Socket resistance

Where piles are driven or more likely drilled some distance into rock, an allowance
can be made for the resistance along the shaft. Socket resistance depends on
many factors such as the strength of the rock, socket roughness, interface bond,
socket contamination, weak rock smear presence etc. The following methods
relate to estimation of shaft resistance.

6.2.1 Empirical methods

On the basis of a comparison of achieved shaft resistance and unconfined

compressive strength of intact rock samples, the following allowable shaft
resistance values are recommended:

𝑓 0.05𝑞

where fs(all) is the allowable unit shaft resistance

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.43
Topic 3A: Axial Capacity of Piles

This criterion should be used with caution for highly fractured rocks, for which
Poulos recommends a maximum allowable shaft resistance of 75 to 150 kPa.

Poulos also suggests a reduction in the recommendation of 30% for piles in uplift.

6.2.2 Kulhawy’s method

On the basis of an analysis of load test data from a series of international

databases, Kulhawy recommends the following value of ultimate shaft resistance
(fsu):
.
𝑓 𝜓 𝑐 .𝑝
where fsu is the ultimate unit shaft resistance
cu = qu/2=UCS/2
pa = atmospheric pressure = 100 kPa
 = 1.0 to 3.0, and averages 2.0

6.2.3 Other notable methods

Bell and Robinson (2012) provided a table as given below, which summarises a
number of methods for the estimation of ultimate unit shaft resistance (fsu) of
bored pile based on average uniaxial comptessive strength of rock (quc) in MPa.

Horvath (1978) fsu=(quc)0.5

Horvath & Kenney (1979) fsu=0.20 to 0.25(quc)0.5

Meigh & Wolski (1979) fsu=(quc)0.6

Rowe & Armitage (1987) fsu=(quc)0.5

Rosenberg & Journeaux (1976) fsu=(quc)0.515

Williams & Pells (1981) fsu=(quc)

Williams & Pells (1981) method, widely used in Australia, requires values of rock
socket reduction factor () and rock mass correction factor () can be found using
the figures below if the uniaxial compressive strength of rock (quc) and mass factor
(j) are known from the laboratory and field investigation. Rock socket reduction
factor () for Horvath (1978) & Rosenberg & Journeaux (1976) methods can also
be found in this figure which are lower than the William & Pells values who
included the rock mass correction factor in their method.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.44
Topic 3A: Axial Capacity of Piles

6.2.4 Pells method

Pells (1999) presented an elastic design method for the rock socketed pile based
on the intial work of Pells & Turner (1979). In this method, the compatibility of
deformation between the shaft and base was considered to develop a base load-
sharing chart incorporating the pile-rock stiffness ratio and the socket geometry.
The ultimate limit state and serviceability loads (AS1170) are required in addition
to the settlement limit for this method. A step by step design procedure for full
sockets in uniform rock was provided in Pells (1999). If non-uniform rocks are
encountered and pile shaft slip is considered then Pells suggested using of Rowe
& Armitage (1984) as illustrated in Pells (1999). Readers are advised to consult
Pells (1999), Pells & Turner (1979) and Rowe & Armitage (1986) for an in-depth
understanding of this elastic design method.

The settlement of the pile at the socket level can be estimated using the elastic
equation as:
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.45
Topic 3A: Axial Capacity of Piles

𝛿 𝐼 where, Pt=serviceability load, Er=rock mass modulus (apply

appropriate reduction factor- use 0.75 if the rock mass modulus is measured
based on pressuremeter test and 0.5 if the mass modulus is estimated based on
correlation such as RMR, RQD), D=socket diameter, I=settlement influence factor
(use chart of Pells for known L/D and Ep/Er ratios).

6.2.5 Monash method

The Monash method has been developed by Seidel (1993). The method is
incorporated into the Windows package “Rocket”, which are used in Melbourne
for designing rock socket piles in mudstone rock.

Notes: Most of the methods presented above assumes the socket is clean and
the base is free from debris. If construction methods are poor and base cleanness
can be ensured then the pile capacity is to be determined based on the socket
resistance only with appropriate factor of safety-often this is expensive. Sockets
are also roughened to ensure a ductile pile-rock interface failure. Different
roughness categories used in construction can be found in Pells (1999). In large
projects, where the accuracy of pile capacity estimation is important, a finite
element modelling should be undertaken.

7. Negative shaft resistance

When piles are placed in a settling ground (soft underconsolidated soil/loose fill),
the soil around the pile moves down relative to the pile thus will induce a
downward axial force (known as negative skin friction or drag force) on the pile.
Poulos (2008) pointed out that this additional downward axial force will have no
impact on the ultimate limit state (ULS) analysis as the plunging failure of pile
requires the pile to move down relative to the suppounding ground. Therefore,
the negative skin friction will not generally reduce the ultimate geotechnical
capacity of the pile. However, Tomlinson & Woodward (2008) suggest to ignore
the contribution to the support provided by friction over the length of a pile in soft
clay.

The effect of the negative skin friction on the pile will cause additional settlement
and this has to be accounted for the serviceability limit state (SLS) design.

Poulos (2008) proposed a method of estimating pile head settlement under a

working load for two layers of soils (top layer is settling) and the bottom is stable
layer (ie. provides positive shaft resitance). For the calculation of the downdrag
force the position of the neutral plane (ie. settlement of pile=soil settlement) can
be considered at the bottom of the top layer. The settlement Sw of the pile head
at the working or serviceability load can be estimated as follows (Poulos, 2008).

Sw = S1 + S2 + S3

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.46
Topic 3A: Axial Capacity of Piles

Where, S1=elastic compression of the portion of the pile shaft in the settling zone
due to the applied load on the pile head (Pw)

S2=elastic compression of the portion of the pile shaft in the settling zone due to
the induced drag forces (Pneg) that can be calculated based on the piles in drained
soil concept (i.e., shaft resistabce piles in granular soil).

S3=settlement of the portion of the pile in the stable zone.

S1 and S2 can be computed from simple column compression theory and S3 can
be computed adequately using the elastic theory for the length of the pile
embedded in the stable zone due to applied pile head load (Pw) plus the maximum
drag force in the pile (PNmax).

Poulos also suggested an alternative method to avoid the pile settlement

continually as the ground settles by satisfying the following criteria:

𝑅 𝐹𝑆 𝑃 𝑃 [in terms of Factor of Safety]

where Rug2=ultimate geotechnical capacity of the pile in the stable zone below the
depth of soil settlement; FS2= Factor of safety for the portion of the pile in the
stable zone ≥ 1.25. Readers are advised to refer to Poulos (2008) for further
details.

8. Axial capacity of pile groups

In the previous sections, the axial capacity of a single pile was discussed. Single
piles can, for example, be used to support isolated column loads or under legs of
transmission structures. However, where column loads are particularly high,
groups of single piles may be needed. Structures with raft foundations may also
be supported on arrays of multiple piles, or pile groups.

Each individual pile is supported by the surrounding and underlying soil.

Conversely, the pile imposes a region of stress influence on the soil, which is
greatest immediately adjacent to the pile, and decreases with increasing distance
from the pile. If multiple piles are used in a group, the regions of stress influence
overlap, and the capacity of the soil to support the piles may be reduced.

In addition to the effect of stress overlap, the capacity of a pile may be influenced
by the installation of neighbouring piles. Installation of displacement piles will tend
to increase soil density and lateral stresses locally around a pile. Installation of a
non-displacement pile may have the opposite effect. However, such broad
generalizations cannot be relied upon. For instance, driving piles into sedimentary
rocks can sometimes lead to significant loss of end-bearing known as relaxation.

In any event, the pile group capacity may be less than the sum of the capacities
of the individual piles which make up the group

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.47
Topic 3A: Axial Capacity of Piles

Pug  n Pup

where, Pug is the ultimate capacity of the pile group;

Pup is the ultimate capacity of an individual pile, and
n is the number of piles in the pile group.

It is common to not allow for any increase in capacity due to densification effects.
However, pile group capacity losses are an effect which engineers must be
careful to account for. Pile group capacity loss is by convention calculated using
a pile group efficiency factor, .

Pug   n Pup

The efficiency can vary considerably. Major factors affecting efficiency are soil
type, pile group size and the ratio of pile diameter to spacing.

9. Pile groups in Clay

For pile groups in clay, it appears both from the results of laboratory and full scale
testing that the end bearing component of resistance is largely unaffected by
group action with the result that the efficiency factor for this component is
effectively one, i.e.

 end ,clay  10
.

By contrast, the side resistance component can be significantly reduced by the

proximity of other piles, depending on whether the piles are free standing or
capped at the ground surface. The spacing/diameter ratio of the piles has an
important influence on efficiency - the efficiency factor progressively reduces as
piles in the group are more closely spaced.

In practice, since piles in clay usually derive most of their capacity through shaft
resistance, it is normal to apply the efficiency factor to the total pile resistance
including both shaft and base resistances, despite the foregoing efficiency factor
of 1.0 for base resistance.

9.1 Free standing pile groups

A free standing pile-group is one in which the piles either extend above the soil
surface to a pile cap clear of the founding layer or one in which the pile cap rests
on ground that cannot be relied on to support any load. There are a number of
methods by which the capacity of such a group can be estimated.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.48
Topic 3A: Axial Capacity of Piles

9.1.1 Feld Rule

The Feld rule is a simple empirical method for estimating an efficiency factor. The
calculated capacity of each individual pile is reduced by one sixteenth for each
neighbouring pile. For example, consider a 3 x 4 pile group as shown below.

A B B A

B C C B

A B B A

In this group, there are three types of pile. Type A (4 No.) is surrounded by 3
neighbours, type B (6 No.) by 5 neighbours, and type C (2 No.) by 8 neighbours.
It follows that the efficiency of the pile group is given by:

  12 16 16 16 
1 4  13  6  11  2  8  0.70

It should be noted that this method does not take into account the spacing of the
piles. This is clearly a deficiency of this method.

9.1.2 Converse-Labarre Formula

The Converse-Labarre formula for pile group efficiency is as follows:

  (n  1)m  (m  1)n 
  1
90  mn 

where m is the number of rows of piles,

n is the number of piles in each row, and
d 
  tan 1  
s 
and s is the centre to centre spacing of the piles, and
d is the pile diameter or width.

Using the example given above for s/d ratios from 2 to 10, the efficiency factor
varies as follows:

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.49
Topic 3A: Axial Capacity of Piles

s/d d/s 
2 0.50 0.58
3 0.33 0.71
4 0.25 0.78
5 0.20 0.82
6 0.17 0.85
8 0.13 0.89
10 0.10 0.91

The Converse-Labarre gives a much more realistic prediction of group behaviour

with the efficiency factor approaching one for large pile spacings (i.e. pile
interaction becomes minimal).

9.1.3 Block Failure

An alternative method to computing an efficiency factor is to compare the sum of

the individual pile capacities with the capacity of a block defined by the outer
perimeter of the group. This method was first proposed by Terzaghi and Peck
(1948). The group capacity is assumed to be the lesser of the two capacities.

The pile group is modelled as a shallow footing capacity, however, the footing
capacity, PBL, includes the cohesion at the side, which is not normally taken into
account for shallow footings, i.e.

PBL  BLc b N c  2(B  L )Dc s

where B is the width of the block,

L is the length of the block,
Nc is the relevant N factor allowing for shape and depth,
cb is the representative cohesion at the pile base,
cs is the representative cohesion at the pile shaft, and
D is the depth of the piles.

Model pile tests conducted by Whitaker (1957) appear to confirm that as the pile
spacing decreases, the failure mechanism changes from one of individual pile
failure with slip lines developing around individual piles to one of block failure with
the slip lines around the perimeter of the pile group. However, this method
suggests that there is a critical spacing at which the failure criterion changes
abruptly from one to the other. As this does not appear to be the case it is
suggested that an empirical relationship can be used to give a more realistic
transition from one failure mode to the other. This relationship is given by
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3A.50
Topic 3A: Axial Capacity of Piles

1 1 1
2
 2 2
 2
Pug n Pup PBL

This same relationship can be rewritten to give an efficiency of

1 n 2Pup
2
 1
2 2
PBL

Poulos and Davis (1980) provided an example of the application of this approach,
where the ultimate load of a pile group of specific dimensions is related to the
number of piles in the group. This is shown below. It can be seen that there is a
general transition between single
pile behaviour and block action. 5m
Furthermore as would be expected,
once block action begins to 5m
dominate, there is little advantage in
increasing the number of piles in the
group. In this particular case,
increasing the number of piles Clay
L = 20m d = 0.3m
beyond about 60 does little to c u = 50 kPa

increase the capacity of the pile

group.
Limit for block failure
Ultimate group load ( M N )

30
Isolated
single pile
The general applicability of this failure
empirical transition approach was
examined by Poulos and Davis in 20
connection with the model pile tests
conducted by Whitaker. Some of
the comparisons are shown in the 10
following two figures for variations in
pile numbers and pile length to
diameter ratios. These figures
suggest that higher efficiencies can 0 5 10 15
be achieved for greater pile No. of piles
spacings, a smaller number of piles
in a group and for lower length to
diameter ratios.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.51
Topic 3A: Axial Capacity of Piles

1.0

0.6

1.0 1.0 1.0 1.0
0.2 52 group 72 group 92 group

 1 2 3 4 1 2 3 4 1  2 3 4
s s s
d d d
0.5 0.5 0.5 0.5
0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8
s s s s
d d d d
L L L L
(a) d = 12 (b) d = 24 (c) d = 36 (d) d = 48

9.2 Capped pile groups

This type of pile group has a rigid pile cap which bears directly on soil which
provides reliable support. In this case, all piles must operate as a group because
of the rigid boundary condition at the pile heads.

Bc Referring to the schematic, the ultimate load

Lc capacity is the lesser of

(i) the sum of the capacity of the block

containing the piles, PBL, plus the capacity of
that area of the cap lying outside the block,
Pcap. The second component is estimated as:

Pcap  cN c B c Lc  BL 
L

B where Nc is the relevant value of bearing

capacity factor allowing for depth and shape,
and c is the cohesion relevant to the bearing of the cap.
(ii) the sum of the individual pile capacities, nPu, plus the bearing capacity of the
pile cap on the bearing stratum, Pcap. The individual capacity of the piles should
be calculated according to the single pile axial capacity methods. In this case, the
bearing capacity of the pile cap is given by


Pcap  cN c Bc Lc  nAp 

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.52
Topic 3A: Axial Capacity of Single Piles

where, in addition to previous definitions,

Ap is the cross-sectional area of each pile.

Clearly, when the piles are relatively close together, the first mode will govern the
capacity of the group, but as the spacings of individual piles increases, the second
mechanism will predominate. This general behaviour for capped pile groups is also
confirmed by the model pile test results of Whitaker (1960) as shown in the following
figure.

For closely spaced piles, both free

1.0
standing and capped groups give
very similar responses. However,
0.9
for the free standing pile groups, 3
2

once individual behaviour takes 3

2
0.8
over for greater spacings, the rate
Efficiency , 

of increase of efficiency
0.7
decreases. For the capped 5 2,7 2 ,and 92
groups, the rate of increase of
0.6
efficiency continues as before 2
5
because of the larger pile cap Tests on freestanding
0.5 groups
influence. 72 Tests on piled
foundations
0.4 Calculated for piled
2
foundations, assuming
9 block failure
0.3
1 2 3 4
Spacing factor, s/d

10. Pile groups in Granular Soils

There is very little detailed information in the literature concerning the capacity of
pile groups in sand. For driven piles, it seems fairly certain that for sand that is less
than dense, the efficiency of a group will generally be greater than unity. It would
appear from various model tests (e.g. Vesic, 1969), that the maximum efficiency
ranges from about 1.3 to 2 at spacings in the general range from 2 to 3 pile widths.

For dense and very dense sand, the driving action could loosen the sand, but as it
is virtually impossible to drive piles through dense sand, this situation is unlikely to
have any practical significance. In the case of bored piles in granular soils, the shaft
capacity does not generally dominate design. Therefore, when considering this
relatively small component of total resistance, the use of an efficiency factor of one
should not cause any major overestimation of capacity.

For end bearing in piles in granular soils, it seems again that an efficiency factor of
unity should be used for both driven and bored piles. However, it must be stressed
in the case of bored piles, that some care should be taken in assessing individual
pile capacity so that any loosening during pile formation is allowed for.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.53
Topic 3A: Axial Capacity of Single Piles

It may be worth pointing out that with piles which are primarily end bearing (which
generally means piles founded in sand, gravel, rock, etc. although it is possible for
this to occur with piles in clay), it is important to check that there is a sufficient
thickness of the bearing stratum under the pile bases so that the bearing capacity is
not reduced significantly.

The underlying softer layer in the following schematic would not affect the capacity
of a single pile, but would be within the zone of influence of the pile group indicated.

Bearing Layer

Softer / weaker Layer

It is important to check that the bearing capacity of the underlying layer is not
exceeded so that the group does not punch through the bearing layer as illustrated
below. Shallow footing analyses would be applied.

Test pile

Fill or
weak soil

Com pact
stratum

Soft com pressible clay

(a) (b)

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3A.54
Topic 3A: Axial Capacity of Single Piles

11. References

Bell, A. & Robinson, C. (2012). Chapter 54-Single piles. ICE manual of Geotechnical
Engineering. ICE publishing, London, doi: 10.1680/moge.57098.0803

Bowles, J. (1997). Foundation analysis and design. 5th ed. McGraw-Hill.

Das, B.M. & Sivakugan, N. (2019). Principles of Foundation Engineering. 9th ed.
Cengage, Boston, USA.

Haberfield, C.M. & Lochaden, A.L.E. (2019). Analysis and design of axially loaded
piles in rock. Journal of Rock Mechanics and Geotechnical Engineering, 11(3), 535-
548.

Kulhawy, F. (1984). Limiting tip and side resistance: fact or fallacy? Analysis and
design of pile foundations. Proc. ASCE Geotech. [Link]. Symposium. Oct., 1984.
Ed. J.R. Meyer : 80 - 98.

Pells, PnJ (1999). State of practice for socketed piles in rocks. Proc. 8th ANZ Conf.,
Hobart, 1, 307-327.

Pells & Turner (1979). Elastic solutions for the design and analysis of rock-socketed
piles. Can. Geotech. J.

Poulos, H.G. & Davis, E.H. (1980). Pile foundation analysis and design. Wiley.

Poulos, H.G. (1989). Pile behaviour-theory and application. Geotechnique, 39(3),

365-415.

Poulos, H.G. (2008). A practical design approach for piles with negative friction. ICE
Proc. Geotechnical Engineering, 6, 19-27.

Tomlinson, M. & Woodward, J. (2008). Pile design & construction practice (5 ed.).
Taylor & Francis, NY, USA.

Rowe, K. R. & Armitage, H.H. (1987). A design method for drilled piers in soft rock.
Can. Geotech. J., 24, 126- 142.

Semple, R. M. & Rigden, W. J. (1984). Shaft capacity of driven pipe piles in clay,
Symposium on Analysis and Design of Pile Foundations, American Society of Civil
Engineers, San Francisco, 59–79 (cited from Tomlinson & Woodward, 2008).

Zhang, L. (2010). Prediction of end-bearing capacity of rock socketed shafts

considering rock quality designation (RQD). Can. Geotech. J., 47(10), 1071-
1084, [Link]

N.B. Not all references are listed here. Readers are suggested to refer to standard
texts listed above for all other references not listed here.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.1
Topic 3B: Settlelment of Piles

Topic 3B: SETTLEMENT OF PILES

TABLE OF CONTENTS

1. Settlement of single piles................................................................ 2

1.1 Load Transfer Method ....................................................................... 3
1.2 Elastic Analytical Methods ................................................................. 6
1.3 Load at the pile base ......................................................................... 7
1.3.1 Floating pile ............................................................................................... 8
1.3.2 Pile bearing on a stiffer layer .................................................................... 9
1.4 Settlement of a floating pile.............................................................. 11
1.5 Settlement of a pile bearing on a stiffer layer ................................... 13
1.6 Influence of pile slip ......................................................................... 15
1.7 Prediction of load settlement curve .................................................. 16
2. Evaluation of Young’s modulus ................................................... 19
2.1 Typical Values of Young’s Modulus of Drained Soil ......................... 21
2.1.1 Clay soils ................................................................................................. 21
2.1.2 Non-cohesive soils (driven piles) ............................................................ 22
3. Numerical Methods ........................................................................ 22
4. Settlement of group piles .............................................................. 23
4.1 Empirical Methods ........................................................................... 23
4.2 Shallow footing analogy ................................................................... 24
4.3 Elastic Analysis Methods ................................................................. 24
4.3.1 Methods of Analysis ................................................................................ 25
4.3.2 Floating pile groups ................................................................................. 26
4.3.3 End bearing pile groups .......................................................................... 30
4.4 Numerical Methods .......................................................................... 30

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.2
Topic 3B: Settlelment of Piles

1. Settlement of single piles

Until relatively recently, the settlement of a single pile was calculated either analytically
based on many simplifying assumptions or on an empirical basis through correlations
with other pile tests in similar situations. Of course, it may be noted that in many
situations, the settlement was not calculated at all but assumed to be satisfactory if
the load did not exceed one third of the ultimate load.

Of the analytical methods, one of the more popular was to assume that for an end
bearing pile, the pile had its entire load concentrated on its base and it could be
analysed in the same way as a shallow foundation. For a friction pile, the same
assumption was made but the equivalent base was assumed to be founded at a depth,
which was two thirds of the length of the pile. Clearly, these assumptions oversimplify
a much more complex situation and cannot be expected to yield accurate predictions
of pile settlement.

There are many examples of the empirical approach including Meyerhof (1959), who,
for piles in clay, suggested that for loads not exceeding one third of the ultimate load,
the settlement is given by

db

30F
where db is the diameter of the base of the pile, and
F is a factor of safety

For piles in sand, Focht (1967) suggested

 MR col

where col is the compression of the pile as a free standing column, and
MR is the movement ratio.

For short piles rigid piles of col < 8 mm, MR = 1. For long compressible piles for
col > 8 mm, MR = 0.5.

However, with the advent of computers, a number of more sophisticated analyses

were developed. These methods permit a far more realistic assessment of pile
settlement because of the incorporation of the many factors that influenced pile
settlement. These various computer based methods fall into three broad categories as
follows:
1. Load transfer methods
2. Elastic analytical methods
3. Numerical methods.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.3
Topic 3B: Settlelment of Piles

As it is nearly always one of these three methods which is employed to estimate the
settlement of a single pile, each of these three methods will be considered in detail
below.

1.1 Load Transfer Method

The load transfer method was originally suggested by Seed & Reese (1957) with
further refinements by Coyle & Reese (1966). The method is basically an iterative
technique, which is used to calculate the load/displacement characteristics at the head
of the pile and, thereby allow the construction of a full load/settlement curve.

The method requires the pile to be divided into a number of segments, which are
assumed to be connected by springs as shown. A small displacement is then applied
to the pile base, and by calculating the forces and displacements for each segment
progressively up the pile shaft, the load at the head of the pile can be determined along
with its vertical displacement. This process is repeated for a number of base
displacements until a sufficient range of pile loads and displacements are obtained to
construct a complete load/settlement curve.

P In order to apply this process, it is necessary

to know the load transfer characteristics of
d each small section of the pile shaft. These
S1 1
z 1 mid characteristics are given by the shear stress
F 1 bot - shear displacement curve at this small
F 1 bot = F 1 top section of interface and are often referred to
F 2 top
as a t-z curve as is shown following. These
curves may be obtained by a number of
S2 2
z 2 mid processes including laboratory tests and
F 2 bot instrumented field tests.

The various steps involved in the analysis

are given here:
L F n-2 top
1. Divide the pile into a number of segments.
S n-2 n-2 Clearly, the accuracy of the technique
z n-2 mid depends on the number of segments
F n-2
bot
F n-2 = F n-1 top chosen. However, it would appear that if the
F n-1 top bot
number is at least 10, a reasonable
accuracy is obtained.
S n-1 n-1
z n-1 mid 2. Considering the bottom segment, n,
F n-1
bot F n-1 = F n top assume a small displacement, znbot, at the
F n top bot
pile base. Based on this small displacement,
Sn
estimate the load, Fnbot, required to
Ln n
z n mid produce this displacement. This can be
F n bot achieved by assuming that the base is rigid
z n bot

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.4
Topic 3B: Settlelment of Piles

and that conventional elastic analyses applies. For example, for a rigid circular
base,

2DEz nbot
Fnbot 
1  2

where D is the diameter of the pile,

E is the Young’s modulus under the base of the pile,
and  is the Poisson’s ratio under the pile base.

It is of course possible to assume that the pile base displacement is zero, which is
appropriate for piles with only small loads or for piles on rigid bearing strata.

3. For the first trial, assume that znmid the displacement at the midpoint of segment
n, is equal to znbot.

4. Using the
adopted t-z
curve, for the
 n mid
value of znmid,
Shear stress , 

read off the

 n-1 mid corresponding
value of nmid.

z n-1 mid z n mid

Shear displacement

5. Calculate the load, Fntop, at the top of segment n as follows:

Fntop  Fnbot  nmid    D  Ln

6. By assuming a linear variation of load distribution along the length of the

segment, the load at the midpoint of the segment is

Fntop  Fnbot
Fnmid 
2
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3B.5
Topic 3B: Settlelment of Piles

7. Calculate the elastic deformation of the bottom half of the segment, n, as follows

Fnmid  Fnbot Ln
z nbot  2 2
AE
where AE is the pile stiffness.

8. Calculate the displacement at the midpoint of the segment n as

z nmid  z nbot  z nbot

9. Compare the value of znmid assumed after step (3) with the value calculated after
step (8). If they are outside a specified tolerance, then assume znmid to have the
value calculated after step (8) and repeat steps (4) to (9) until the tolerance is
satisfied. Normally, this can be achieved after a few iterations.

10. Proceed to the next segment up the pile shaft, segment n -1. Repeat steps (2) to
(10) with the following load and displacement values at the bottom of segment n -1

Fn 1 bot  Fntop

z n 1 bot  z nmid  z ntop

Fntop  Fnmid Ln
z ntop  2 2
AE
11. Continue in this manner up through the pile segments until the load and
displacement at the head of the pile are obtained.

12. Steps (1) to (11) are then repeated for different values of displacement until a
range of load and displacement values are obtained for the head of the pile to
enable a plot of load against settlement.

It is clear from this iterative process that the method is ideally suited for computers.
However, it would be fair to say that the method can only be regarded as approximate
and does suffer from three main disadvantages.

 The first is that the method assumes that there is no significant interaction
between successive pile segments and the soils surrounding these segments.
This is not realistic. However, the model has nevertheless been found to be
reasonably accurate.

 Secondly, the method cannot be used to analyse pile groups.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.6
Topic 3B: Settlelment of Piles

 Finally, the method is only as accurate as the assumed t-z curve, which can be
difficult and costly to obtain. There would seem to be some limited success with
large laboratory shear box methods (Johnston et al., 1987).

Despite the above limitations, Reese, Houdan and Vijayvergiya (1969) have
suggested that for bored piles in clay, the load transfer curve can be generalised by

    
az  a max 2   
 so  so  

where az is the shear resistance at depth z,

amaxis the maximum shear resistance at any depth,
 is the vertical displacement of the pile at depth z,
so is 2 D 
D is the pile diameter,
and  is the average failure strain in percent for uniaxial compression tests
on the clay.

A very useful form of the load transfer method has been developed by Randolph
(1986) and the program is called RATZ. This program is capable of many refined tasks
including the analysis of dynamic loading.

1.2 Elastic Analytical Methods

Probably the most successful forms of analysis for the settlement of piles foundations
is based on methods of elasticity in which Mindlin’s equations for the displacements
within a soil mass caused by loading within the mass are used. Several investigators
have used this approach but perhaps one of the most complete sets of solutions have
been developed by Poulos and Davis (1980). Vesic (1977) also presented a rigorous
method based on elastic analysis.

While there are arguments that pile behaviour is not necessarily totally elastic, the
solutions obtained do seem to capture the essential features of pile settlement and the
results obtained do seem to compare well with piles that have been instrumented. Pile
working loads are generally sufficiently below the pile ultimate capacity to ensure
essentially linear load/settlement response up to these loads. Therefore, the
application of elastic methods can be justified.

1.2.1 Elastic method of Poulos & Davies (1980)

This technique involves the discretisation of the pile into a number of elements. It is
then necessary to obtain mathematical expressions for the vertical displacement of
Department of Civil Engineering, Monash University
Revised: 6/2020
CIV4249: Foundation Engineering 3B.7
Topic 3B: Settlelment of Piles

the pile and the soil at each element in terms of the unknown stresses on the pile. By
applying the compatibility conditions to the pile and the soil, the various equations can
be solved to yield the displacement at any given pile head load. Details of these
analytical methods are discussed in some detail in Poulos and Davis (1980).

One of the strong points of the various analyses that have been carried out is that
design charts for a wide range of pile conditions have been developed using
dimensionless parameters. These have made the analysis of pile settlements
comparatively easy for a wide range of commonly encountered conditions.

In the various analyses that follow, two variables are frequently used. The first is the
area ratio, RA, defined as
Ap
RA 
As
where Ap is the area of the material of the pile section,
and As is the area bounded by the perimeter of the pile.

For a solid concrete pile, Ap = As, and it follows that RA = 1. However, for a steel pipe
pile, Ap < As, and therefore, the area ratio would be much less than one.

The second variable is the pile stiffness factor, K, which is

Ep
K  RA
Es
where Ep is the Young’s modulus of the pile material,
and Es is the Young’s modulus of the soil.

1.2.2 Load at the pile base

It may sometimes be necessary to establish how much of the total pile load is
transferred to the base of the pile. While this is not directly connected to pile
settlement, it is given here as the analytical techniques used are the same as for the
settlement analysis. At working load, many piles carry very little of their load through
base resistance, even though much of their ultimate resistance may be derived though
end bearing. The longer and more flexible the pile, the higher will be the percentage
of load carried by shaft resistance at working load.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.8
Topic 3B: Settlelment of Piles

[Link] Floating pile P

The situation is defined as follows.
The proportion of the total load Soil -
reaching the pile base, , is given by: Young's modulus, E s
L Poisson's ratio,  s
  oCK C  d
h

where o is the proportion on the pile

Pile
base for a rigid pile, CK is a Young's modulus, E p
correction factor for pile
compressibility, and C is a
correction factor for Poisson’s ratio of
Rigid stratum
the soil.
The values of o, CK, and C are given in the following figures.
1.0
1.0
0.8
0.8
0.6 2
5
Values of d b/d 0.6
0.4 10
CK
3
0.4
25
2 Values of L /d
0.2 0.2
50
o 1
0 5
4
10 100 1000 10 10
0.1
K
0.08
1.0
0.06

0.04 0.9

Values of K

0.8
10000
C
0.02 1000
0 10 20 30 40 50
0.7
L/d 100

0.6

0.5
0 0.1 0.2 0.3 0.4 0.5
s

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.9
Topic 3B: Settlelment of Piles

The first of these charts also considers the effect of piles with enlarged bases. From
these figures, it is clear that the percentage of load reaching the pile base increases
considerably with enlarged bases.

[Link] Pile bearing on a stiffer layer

In this case, the percentage of the total

load reaching the pile base is given by: P

  oCK C Cb Pile -

Young's m odulus E p
d
where, in addition to the above L
Soil -
definitions, Young's modulus Es
Poisson's ratio  s
Cb is a correction factor for the stiffness
of the stiffer layer and is given in the Bearing Stratum -
Young's modulus Eb
following figures. Poisson's ratio  b

20
Values of K Values of K
20 000 5000 20 000
5000
1000 1000
10

8
500 500
6
Cb
4

100

100
2

(a) L
d
= 75 (b) L = 50
d

1
1 10 100 1000 1 10 100 1000
Eb Eb
Es Es

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.10
Topic 3B: Settlelment of Piles

10
10 000
8 1000 500
100
6
Values of K

4
20
Values of K Cb
20 000 1000
500
10 2
8 100 L
(d) = 10
d
6
Cb 1
1 10 100 1000
4 Eb
L
Es
(c) = 25
d
Values of K
2 5
500, 1 000, 5 000, 20
4 000
100
3
11 10 100 1000 Cb
Eb 2
Es
L
(e) =5
d

11 10 100 1000
Eb
Es

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.11
Topic 3B: Settlelment of Piles

1.2.3 Settlement of a floating pile

The situation considered is the P

same as illustrated here. The
settlement at the pile head, , is
given by: Soil -
Young's modulus, E s
PI
 L Poisson's ratio,  s
E sd
d
h
I  IoRK RLR 

where d is the pile diameter; Pile

Young's modulus, E p
Io is a settlement
influence factor for a rigid pile and
a soil with  = 0.5;
RK is a correction factor for pile Rigid stratum
compressibility;
RL is a correction factor for a finite thickness of soil between the base of the pile and
a rigid layer; and R is a correction factor for the Poisson’s ratio of the soil.

The values of Io, RK, RL and R are given in the following figures.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.12
Topic 3B: Settlelment of Piles

1.0 3
0.8
100 Values of L d
0.6
50
R
0.4 RK 25

2
Values of db /d
10
1
0.2
2
3
Io 5
2
0.1

0.08 1 1
10 100 1000 10 000
0.06 K

0.04 1.0
For L d = 100
I o = 0.0254
Values of L d
For 3 > d b d > 1 50
0.8
0.02 25
0 10 20 30 40 50 Rh 10
5
L 0.6
d
2

0.4 1

0.2

0
0 1 h 2
L 0.5 L 0
h

1 .0 0

K = 100
0 .9 0
500
R  2000
1000

0 .7 5
0 0 .1 0 .2 0 .3 0 .4 0 .5
 s

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.13
Topic 3B: Settlelment of Piles

1.2.4 Settlement of a pile bearing on a stiffer layer

The settlement of a pile bearing on a stiffer layer is computed as follows, i.e.

PI P

E sd

Pile -
where, I  IoRK RbR 
Young's modulus Ep
d
In addition to the definitions
given above for Io, RK and R, L
Soil -
Rb is a correction factor for the Young's modulus Es
stiffness of the stiffer layer and is Poisson's ratio  s
given in the following set of
charts.
Bearing Stratum -
Young's modulus Eb
Poisson's ratio  b

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.14
Topic 3B: Settlelment of Piles

1.0 1.0
100 100
0.8 500 0.8
1000 500
0.6 0.6
Rb > 20 000 5000 Rb > 20 000 1000
0.4 0.4
Values of K Values of K 5000
0.2 0.2
(a) L d = 75 (b) L d = 50
0 0
1 10 E b 100 1000 1 10 E b 100 1000
Es Es
1.0 1.0
100
0.8 0.8
0.6 0.6
Rb 500
0.4
Rb 100
0.4 1000
1000 > 10 000
0.2 Values of K 0.2 L
500
L
(c) d = 25 > 2000 (d) d = 10
0 0
1 10 E b 100 1000 1 10 E b 100 1000
Es Es
1.0
Values of K

0.8

0.6
Rb
0.4
(e) L d = 5 100
0.2
5000 500 1000
20 000
0
1 10 Eb 100 1000
Es

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.15
Topic 3B: Settlelment of Piles

This figure shows that for very long compressible piles, the presence of a stiffer layer
at the pile base has little effect.

An alternative method of solution has been proposed by Poulos and Davis to include
the concept of the movement ratio, MR, where

settlement of pile
MR 
elastic shortening of pile

PL
and  MR
EA

1.0 L =5
d
10
25
0.8
50
100

0.6

MR

0.4

PL
 = M
0.2 EP A P R
 s = 0.5
s = 0
0
100 200 500 1000 2000 5000 10 000
K

For a pile bearing on a rigid stratum, this figure gives the relevant values of MR. This
method is although relatively simple but it does not give as general a solution as was
presented earlier.

1.2.5 Influence of pile slip

Poulos and Davis (1980) provide some illustrations of how slip between the pile and
the surrounding soil can influence pile settlement when elastic analyses are employed.
For the purpose of illustration presented here, only the relatively simple case of a
floating pile in a cohesive soil will be considered. A slip factor, Ms, is defined so that

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.16
Topic 3B: Settlelment of Piles

elastic settlement of pile

Ms 
total settlement of pile

1.0 1.0
1
> 2000
2
0.8 0.8
3
100
0.6 0.6
K = oo Values of K
Ms Ms 10
vs = 0.5
0.4 0.4 L = 0.5
d
L ca s = 0.5
Curve d cu ca = 1
100 1 cu
0.2 1 0.2
2 25 1
3 25 0.4
0 0
2.25 2.0 1.75 1.5 1.25 1.0 6 5 4 3 2 1
Factor of Safety Factor of Safety

These two figures show the effect of pile slip for a constant adhesion along the shaft
of the pile in terms of factor of safety for a few variations in L/d and ca/cu, and for
constant L/d and ca/cu ratios. It would appear from this that provided the pile is
relatively slender (L/d > 25), the influence of L/d is not very pronounced. It would
appear that to working loads of about half ultimate loads, the pile response is primarily
elastic therefore providing some justification of the use of elastic methods. As the ratio
ca/cu decreases, the effect of slip becomes more pronounced. A further point, which
is evident, is that the lower the value of pile stiffness, K, the more significant is the
effect of soil slip.

Based on these observations, it would appear that long steel piles in soft clay are
unlikely to significantly be influenced by soil slip and therefore are ideal for analysis by
elastic methods. On the other hand, relatively short, large diameter concrete piles in
stiff clay will be influenced by side slip as the side shear resistance peaks long before
the base resistance (Whitaker and Cooke, 1966). Therefore, for these pile types,
purely elastic analyses to calculate settlements at working loads may not be
appropriate. Clearly, for these same pile types cast in soft rock where the pile stiffness
is even lower, the load/deformation curve departs linearity even earlier (Williams, et
al. 1980). For such piles, an alternative method of calculating settlement after the
development of significant slip is presented below.

1.2.6 Prediction of load settlement curve

As was pointed out in the previous section, slip on the pile-soil interface has little
influence on piles, which are relatively long, with a fairly stiff section and in weaker
soils. It follows that the load/settlement curves for these piles will be relatively linear to

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.17
Topic 3B: Settlelment of Piles

working loads of up to about half the ultimate load. Therefore, a reasonable

approximation for the load/settlement response for these piles may be obtained by
using elastic methods. Indeed, provided the pile is effectively non-compressible,
elastic methods right up to failure may be reasonable.

For concrete piles with low L/d ratios in stiff clay however, the shaft interface usually
experiences full slip long before the base resistance reaches its ultimate value. The
following method gives a means of estimating the load/settlement curve.

The first step involves calculating the settlement experience by the shaft. The load on
the shaft is estimated from

Ps  P (1  )

where  is estimated from the previous methods for floating piles or for piles bearing
on a stiffer layer.

If it assumed that the settlement is linear up to the full capacity of the shaft, Psu, then
the settlement of the pile head with respect to Ps can be calculated from

where

I is determined from previous elastic methods

and Es is the Young’s modulus along the pile shaft.

The load carried by the base, Pb, is calculated from the total load as

Pb  P

Again, by assuming that the displacement of the base alone up to failure is linear, the
settlement of the base is given by

I Pb

E sd b

Where, Es is again the Young’ modulus along the pile shaft. It should be noted that if
the modulus below the base is different from around the pile shaft, then this is allowed
for by use of the appropriate elastic model.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.18
Topic 3B: Settlelment of Piles

However, this equation does not include the compression of the pile shaft for loads
greater than Psu. For a shaft load of Psu, the load at the head of the pile is given by
.

Ps 
and the proportion of this which reaches the pile base is given by
1 

Therefore, the additional compression of the pile shaft for loads on the shaft greater
than Psu, is given by

 P  L 
   Pb  su  
 1    AE 

It follows that the base load/settlement relationship is

I Pb  P  L 
   Pb  su 
E s d   (1  ) AE  

With the above equations, it is possible to construct an approximate load/settlement

curve. The first step is to calculate the ultimate load for the shaft, Psu. Then, calculate
the head load that gives full shaft slip. This is given by

Psu
Pys 
1 

The settlement at which this occurs is

I
 ys  Pys
E sd

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.19
Topic 3B: Settlelment of Piles

Total
Pu

Base
Load

Pbu

Pys
Shaft
P su

ys u
Settlement

It is now possible to construct the load/settlement curve for the shaft as shown in this
figure, as well as the first leg of the full load/settlement curve.

It is now necessary to calculate the load/settlement curve for the base. This requires
calculation of the ultimate base load, then using

I Pbu  P  L 
u    Pbu  su  
E sd   1    AE 

the settlement of the base, u, at which the base ultimate is reached can be calculated.
This then allows the base load/settlement curve to be constructed as shown above.
By adding the two curves together, the peak for the full load/ settlement curve is at Pu
= Pbu + Psu at a settlement of u.

1.3 Elastic method of Vesic (1977)

Vesic (1977) proposed a semiempirical method based on elastic analysis where the
total elastic pile head settlement (s) has been divided into three components: (i) elastic
shortening of the pile (se1), (ii) settlement (se2) due to the working load at the pile point
(Qwp), and (iii) settelemnt (se3) due to the working load along the pile shaft (Qws).

i.e., se= se1+ se2+ se3

Where, 𝑠

𝑄 𝑊𝑜𝑟𝑘𝑖𝑛𝑔 𝑙𝑜𝑎𝑑 𝑎𝑡 𝑡ℎ𝑒 𝑝𝑖𝑙𝑒 𝑝𝑜𝑖𝑛𝑡

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.20
Topic 3B: Settlelment of Piles

𝑄 𝑊𝑜𝑟𝑘𝑖𝑛𝑔 𝑙𝑜𝑎𝑑 𝑎𝑙𝑜𝑛𝑔 𝑡ℎ𝑒 𝑝𝑖𝑙𝑒 𝑠ℎ𝑎𝑓𝑡

𝐿 𝐿𝑒𝑛𝑔𝑡ℎ 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑖𝑙𝑒

𝐴 𝐶𝑟𝑜𝑠𝑠 𝑠𝑒𝑐𝑡𝑖𝑜𝑛𝑎𝑙 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑖𝑙𝑒

𝐸 𝑀𝑜𝑑𝑢𝑙𝑢𝑠 𝑜𝑓 𝑒𝑙𝑎𝑠𝑡𝑖𝑐𝑖𝑡𝑦 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑖𝑙𝑒 𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙

𝜉 𝐴 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 (𝑓𝑜𝑟 𝑢𝑛𝑖𝑓𝑜𝑟𝑚 𝑜𝑟 𝑝𝑎𝑟𝑎𝑏𝑜𝑙𝑖𝑐 𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛 𝑜𝑓 𝑠ℎ𝑎𝑓𝑡, 𝜉

0.5; 𝑙𝑖𝑛𝑒𝑎𝑟 𝑖𝑛𝑐𝑟𝑒𝑎𝑠𝑒 𝑜𝑟 𝑑𝑒𝑐𝑟𝑒𝑎𝑠𝑒 𝑤𝑖𝑡ℎ 𝑙𝑒𝑛𝑔𝑡ℎ, 𝜉 0.67)
The above equation for se1 takes into account the variation of shaft resistance along
the pile on the elastic shortening of the pile itself under the compressive load.

The settlement of the pile point (se2) due to the working load (Qwp) can be calculated
from the elastic equation presented for the shallow foundation case previously.

𝑠 1 𝜇 𝐼 where, D=width or diameter of the pile; s=Poisson’s ratio

of the soil; Iwp=settelemnt influence factor≈0.85; Es=modulus of elasticity of soil at or
below pile point.

A semi-empirical method for calculating se2 was also proposed by Vesic (1977) as:

𝑠 where, qp=ultimate point resistance of the pile, Cp=empirical coefficient

(values given in table below).

Soil type Cp (typical)

Driven pile Bored pile

Sand (dense to loose) 0.02-0.04 0.09-0.18

Clay (stiff to soft) 0.02-0.03 0.03-0.06

Silt (dense to loose) 0.03-0.05 0.09-0.12

The elastic settlement of the pile (se3) due to the average working load along the shaft
was given by:

𝑠 1 𝜇 𝐼 where, D=width or diameter of the pile; s=Poisson’s

ratio of the soil; Iws=settelemnt influence factor (Vesic, 1977)=2+0.35(L/D)0.5;
Es=modulus of elasticity of soil at or below pile point; p=perimeter of the pile

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.21
Topic 3B: Settlelment of Piles

A semi-empirical method for calculating se3 was also proposed by Vesic (1977) as:

𝑠 where, qp=ultimate point resistance of the pile, Cs=an empirical

constant= 0.93 0.16 𝐿/𝐷 𝐶 .

If single piles are placed in overconsolidated clay then the consolidation component
of the settlement could be small and often neglected. However, for group piles in clay,
the consolidation settlement of the group is to be estimated separately. This can be
done using the consildation settlement concepts discussed previously for shallow
footing.

2. Evaluation of Young’s modulus

In the foregoing sections, it is clear that the critical soil parameter required for the
estimation of pile settlement is the modulus of the soil. The accuracy with which this is
achieved will determine how accurate the predictions of settlement are. The available
methods for determining modulus are many and varied ranging from small laboratory
sized tests on samples obtained as undisturbed as possible from the field to large
scale in-situ tests at the site on which construction is to take place. Clearly, the choice
of which method should be used in a particular situation will be governed by a number
of variables, one of which is cost. Some general discussions on the relative methods
of a number of approaches are presented in Johnston (1983).

2.1 Typical Values of Young’s Modulus (drained soil)

2.1.1 Clay soils

Poulos & Davies (1980) provided back

calculated values of drained modulus (E’) of soils
from the pile load tests. E values were correlated
with undrained shear strength of clay for driven
and bored piles. The modulus value for driven
pile in soft to medium clay is higher than the
bored pile where post driven reconsolidation
makes the clay stronger. However, for stif clay
this is opposite as the “whip” effect during pile
driving causes a reduction of the clay modulus.
The undrained modulus (Eu) can be estimated
using the following relationship in the absence of
any test data.

3𝐸′
𝐸
2 1 𝜇

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.22
Topic 3B: Settlelment of Piles

2.1.2 Non-cohesive soils (driven piles)

Soil Young’s modulus Es MPa

Loose sand 27.5-55

Medium sand 55- 70
Dense sand 70-110
Medium gravel 200

It is generally considered that moduli obtained from small-scale tests in the laboratory
are not advisable for soils, which exhibit considerable disturbance on sampling and
soils that contain significant fabric can cause major scale effects. This basically means
that all granular soils and cohesive soils containing fissures, joints, lenses, inclusions,
etc. are not really suitable for small-scale laboratory tests. These types of soil may be
better to test in-situ. The best form of test is a pile-loading test, which is back-analysed
by elastic methods to obtain a modulus. This modulus can then be used to predict the
behaviour of piles of different sizes with a reasonable degree of accuracy. These tests
are however, not done without some cost.

Also, Young’s modulus of sands for bored piles is very difficult to estimate due to the
effect of installation method affecting the density of sand.

2.2 Poisson’s ratio (Drained soil)

2.2.1 Clayey soil

Stiff (OC) clay ’=0.1-0.2

Medium clay ’=0.2-0.35

Soft NC clay ’=0.35-0.45

2.2.2 Sandy soil

Poisson’s ratio of sands can vary between 0.25 and 0.35. An average value of 0.30
can be used when no test results are available.

3. Numerical Methods

Numerical techniques such as the finite element method and the boundary element
method have been developed for the analyses of pile behaviour by many investigators.
While these methods are relatively complex to use, they do provide facilities for the
realistic analyses of much more complex problems than can be handled by the load
transfer method or by methods based on elastic theory. However, as even a partial
description of the techniques that are available for piles would require significantly

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.23
Topic 3B: Settlelment of Piles

more space than is available, it is suggested that more specific texts on the subject
are consulted.

4. Settlement of group piles

Individual piles are surrounded by a zone of influence - stressed soil, which accepts
the loads transferred from the pile shaft and base. If the pile load is sufficiently small,
it can be assumed that there is no slip between pile and soil, and that the settlement
can be computed using elastic analysis methods.

As discussed with reference to the axial capacity of piles, when the piles are placed in
a group, the individual zones of influence overlap, and the same soil may be subjected
to stresses from many piles. The net result of this effect is that the capacity of an
individual pile in a group will reduce relative to its single pile capacity (There may be
other factors, such as densification of sands in the group, which may counteract this
effect, and even result in a capacity increase).

For the same reasons, when piles are in a group, the settlement effects of an individual
pile will tend to be additive, and the settlement of a pile group will therefore be greater.
The settlement of group piles depends on such factors as the number of piles in the
group, whether the piles carry their settlement by shaft friction or end bearing, and the
spacing/diameter ratio.

There have been a number of approaches proposed for determination of pile group
settlement, and these will be discussed hereafter. Empirical methods are perhaps the
least reliable, and the shallow footing analogy is overly simplistic. Elastic methods of
analysis are perhaps the most suitable method for the majority of engineering
problems. For more critical structures, finite element or boundary element analyses
may be more appropriate.

It is also important to note that the zone of influence of a pile group will extend
considerably further below the pile bases than the zone of influence of the individual
piles. If conditions below the pile group base improve, the net increase in pile group
settlement will be reduced. However, if a soft layer lies within the extended group zone
of influence, special consideration must be taken to include the increased settlements
due to compression of that layer.

4.1 Empirical Methods

Empirical approaches to pile group settlement analysis are based on the results of a
limited number of pile group load tests, either in field or laboratory situations. The test
data is particularly limited because of the difficulty and cost involved in test loading pile
groups.

To extrapolate these particular results to a general recommendation for a wide range

of pile types and conditions is potentially dangerous because of the limited database
that covers only specific soil conditions. Nevertheless, the literature does contain

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.24
Topic 3B: Settlelment of Piles

empirical relationships such as the one suggested for driven piles in sand by Meyerhof
(1959):

s (5  s3 )
g  p
(1  r1 ) 2

where s is the ratio of pile spacing to pile width, and

r is the numbers of rows in a square group.

4.2 Shallow footing analogy

It is normally assumed that the pile group acts as a block, the dimensions of which are
defined by the perimeter of the group and a depth which is equal to either the depth of
the group for piles which are principally end bearing or to two thirds the depth of the
group for friction piles as shown below.

End bearing group Friction group

2 D
3
D

The block is then assumed to behave as a conventional foundation, which is loaded

only at its base and which can be analysed according to the methods described
previously for shallow foundations.

However, this method fails to take into account the way in which piles transfer their
load to the ground, the compressibility of the piles and the very important interaction
between the piles in a group. The fundamental assumption of block behaviour must
be questioned, because as discussed with respect to the axial capacity of piles in a
group, depending on the particular soil conditions and group geometry, the piles may
preferentially fail individually. In such a case, the block analogy would be
inappropriate. This method is at best approximate, and may give misleading
predictions.

4.3 Elastic Analysis Methods

The use of elastic analysis methods for pile group settlement has been aided
significantly by the advent of high-power computers since the 1960s. Although some
of these methods are relatively simple and can be performed manually, the majority

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.25
Topic 3B: Settlelment of Piles

make use of computers, which can perform the numerical iterations necessary to
obtain a converged solution. Poulos and Davis (1980) have obtained solutions for a
vast range of boundary conditions, and have presented the results in dimensionless
chart form for general design use. The use of these charts will be discussed in the
following section.

4.3.1 Methods of Analysis

The methods used for this form of analysis make use of Mindlin's equations for a
loaded and buried cylinder, and a loaded and buried base in an elastic halfspace.
However, for a pile group, it is necessary to consider the interaction between adjacent
piles and this can be done by integration of the Mindlin equations for adjacent piles.
By developing expressions for vertical displacements in terms of the unknown
stresses, and by applying the compatibility conditions, the displacements can be
obtained by solving the resultant equations. The results have initially been expressed
in terms of an interaction factor, , which can be defined as :

additional settlement caused by adjacent pile


settlement of pile under its own load

It should be noted that for all the solutions that are to be considered in this section, the
piles are considered as a free standing group with the pile cap located above the
principal founding layer and not contributing to the overall performance of the group.
The reason for this is simply that with a capped pile group, the interaction between the
piles and the cap acting as a raft is extremely complex and would probably be better
solved using numerical methods.

The following graph shows that as the spacing of the piles increases, so the interaction
factor decreases. Conversely, as the pile stiffness, K, increases so does the
interaction factor. Studies carried out on piles of different L/d ratio shows that
increasing values of L/d will cause the interaction factor to increase.

1.0

L
d = 25
0.8
 s = 0.5
Values of K
0.6
F oo
1000

0.4
500
100
0.2
10

0
0 1 2 3 4 5
s 0.2 0.15 0.1 0.05 0
d d
s

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.26
Topic 3B: Settlelment of Piles

By considering other geometries, it is possible to allow for a stiffer layer underneath

the base of the piles, as well as an increase in pile base diameter. For example, for an
L/d ratio of 25, for K = 0, the following graph gives a correction factor, 𝑁 , which
corrects for the piles being founded in a finite layer of thickness b, underlain by a rigid
layer. The interaction factor, , is then calculated from:

𝛼 𝛼 𝑁

Where, 𝛼 is the interaction factor for floating piles (values of 𝛼 for L/d=10, 50 and
100 can be found in Poulos & Davies, 1980).

1.0 oo
Values of h
5 L
0.8
2.5

0.6
Nh
1.5
0.4
s
L
0.2 h 1.2
L = 25
d
K = oo
s = 0.5
0
0 s 5
d
0.2 d 0
s

The graph shows that as the bases of the piles get closer to the rigid layer, the
interaction factor decreases.

In order to produce solutions that are reasonably easily handled, it is necessary to

make use of a settlement ratio which is defined by

average group settlement

Rs 
settlement of a single pile in the group with the same average load

4.3.2 Floating pile groups

The settlement of a floating pile group in a semi-infinite soil formation can be calculated
from

g  Rs p

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.27
Topic 3B: Settlelment of Piles

The values of Rs are presented in Table 6.2 (p.121) (Poulos & Davies, 1980) as a
function of the L/d and s/d ratios, the number of piles in a group and the pile stiffness,
K. This table shows that the Rs factor increases as the s/d ratio decreases, but it
decreases as the number of piles decreases. As the pile stiffness increases, the value
of Rs also increases. The Rs values given in Table 6.2 are for square pile groups.
However, it should be noted that the shape of the pile group has little influence on the
individual values of Rs, and therefore, this table may be used for other similar shapes
of groups with the same number of piles.

For groups containing more than 16 piles, Poulos and Davis (1980) point out that Rs
varies approximately linearly with the square root of the number of piles in the group.
Therefore, for pile groups containing more than 25 piles, the Rs factor may be
reasonably estimated by selecting R16 and R25 for the same L/d , s/d and K, and
applying the following expression

R s  (R 25  R 16 )( n  5)  R 25

where n is the number of piles in the group.

It may be of some interest to examine the effect of increasing the number of piles in a
group and the effect this has on settlement. To do this it may be better to consider an
alternative reduction factor which is defined as follows

average group settlement

RG 
settlement of a single pile with the same total load

First, consider the situation of increasing the number of piles so that the piles have the
same s/d ratio. Clearly, this means that the width and breadth of the pile group is
increasing. A solution to this problem is shown below for an L/d ratio of 25.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.28
Topic 3B: Settlelment of Piles

1.0
Single pile
0.9 L = 25
d
h = oo
Group reduction factor , R G 0.8 L
Group s = 0.5
0.7
2
0.6
3
2
0.5 2
0.4 5
3x2 2
3
0.3
42
0.2
52
0.1

0
0 1 2 3 4 5 d
s s
d 0.2 0.1 0

It can be seen that such an increase will reduce the settlement of the group with the
load being shared between more piles. However, it should be pointed out that the
settlement reduction seems to become increasingly less effective as the pile numbers
are increased. This is particularly true of piles of relatively close spacing. For example,
consider a pile group of s/d ratio of 3. For a 2 x 2 group, RG is about 0.71. If the
number of piles is increased to give a 3 x 3 group (i.e. increasing the number of piles
from 4 to 9), the value of RG decreases to 0.58. This is not a particularly large
reduction in settlement for over twice as many piles. If the group was further increased
to 5 x 5 piles (giving an increase in the number of piles to 25), the RG value would
decrease to 0.44. It follows that if the number of piles were increased from 4 to 25, the
total settlement would be a little over half the settlement for 4 piles. This seems a fairly
high price to pay. It should also be borne in mind that such an increase in piles for the
same s/d ratio would mean that the pile cap would have to increase dramatically,
which when considering the finite size of a building, may not be at all practical.

What would be more practical, at least as far as construction is concerned, would be

to increase the number of piles supporting a given size of pile cap, i.e. without
increasing the overall breadth of the group of piles. This situation is examined in the
following figure in which the reduction factor RG is related to the ratio of group breadth
to pile diameter.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.29
Topic 3B: Settlelment of Piles

1.0

0.9
L d = 25
0.8 Design curve for h = oo
driven pile groups L
Group reduction factor , RG in sand  s = 0.5
0.7 (Skempton,1953) K = oo
Points used by Skempton
0.6 to derive design curve

0.5
Group
0.4
2
2
2
0.3 3
2
4 2
0.2 5

0.1

0
0 10 20 30 40
Group breadth / d

This figure shows that for a given breadth to diameter ratio, there is little reduction in
settlement beyond a certain number of piles. It would appear then that unless the
original pile spacing is greater than about 6, there is little to be gained by increasing
the number of piles supporting a given pile cap, especially when there are over about
25 piles in the group. This general conclusion seems to hold for a very wide range of
different pile conditions.

Where a rigid layer exists below the pile bases, the group settlement will be reduced
to some extent. The correction factor for settlement for some particular cases is given
in the following figure. It will be seen that the settlement reduces as the depth to the
rigid layer decreases. It also decreases as the size of the pile group increases. More
details of this situation may be found in Poulos (1968).

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.30
Topic 3B: Settlelment of Piles

1.0

22
32
0.8 42
52
2
2
0.6
h  s = 0.5
32
s d = 2.5
42
0.4
52

L = 25
d
0.2 L = 50
d

0
1 h 2
L L
0.5 h 0

4.3.3 End bearing pile groups

The settlement of such a group of piles can be calculated using the following equation
in conjunction with an appropriate table of values of Rs below.

g  Rs p

It will be noted that whereas for friction piles, an increase in pile stiffness leads to an
increase in Rs, for end bearing piles, an increase in stiffness leads to a reduction in
Rs. Where the bearing layer is non-rigid, Mattes and Poulos (1974) give methods of
calculating group settlement.

It should be noted that where the number of piles exceeds 25, the relationship given
previously may be used:

R s  (R 25  R 16 )( n  5)  R 25

4.4 Numerical Methods

Finite element and boundary element analyses are a sophisticated approach to

predicting pile group settlements, particularly for complex boundary conditions. These
analyses may be extremely complex to set up, particularly if 3 dimensional analysis is
required, and axisymmetry does not exist. This type of analysis would only be

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3B.31
Topic 3B: Settlelment of Piles

considered for projects where settlement is extremely critical to the structure, and
where a highly sophisticated analysis can be justified.

It should be noted that the accuracy of these methods is critically dependent on the
failure criterion adopted, and particularly the constitutive laws that govern the pile-
soil interface. The reader is referred to the numerous texts and papers, which deal
specifically with the numerical methods.

References

Das, B.M. & Sivakugan, N. (2019). Principles of foundation engineering. Cingage,

USA.

Poulos, H.G. & Davis, E.H. (1980). Pile foundation analysis and design. Wiley.

Tomlinson, M. & Woodward, J. (2008). Pile design & construction practice. Taylor &
Francis, NY, USA.

Department of Civil Engineering, Monash University

Revised: 6/2020
CIV4249: Foundation Engineering 3C1
Topic 3C: Laterally loaded piles

Topic 3C: LATERALLY LOADED PILES

TABLE OF CONTENTS

1. Introduction ...................................................................................... 2
2. Design methods ............................................................................... 2
2.1 Brinch Hansen’s method .................................................................... 4
2.2 Broms method ................................................................................. 10
2.3 Elastic method ................................................................................. 12
2.4 p-y method for lateral capacity ......................................................... 14
3. Laterally loaded piles in a group .................................................. 18

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C2
Topic 3C: Laterally loaded piles

1. Introduction

In the previous chapter, the axial load capacity of piles was discussed. However, piles
are used to support lateral loads and moments in addition to the axial load. In general,
lateral loads and moments are caused by wind, earthquake, wave and earth
pressures. Similar to axially loaded piles, piles subjected to lateral loads are to be
designed based on both the ultimate limit state and serviceability limit state criteria.

2. Design methods

The lateral load capacity of a pile is derived from the passive resistance of the soil
surrounding it. The distribution of the soil reaction depends on the stiffness of the pile,
the stiffness of the soil and the fixity of the pile. If a pile head is free and the length to
width ratio is below 10-12 then the pile tends to show a rotational failure. The lateral
movement of the pile causes to the mobilisation of passive resistance at the front of
the pile and soil deforms elastically. However, with larger pile head displacements, the
soil in the front fails plastically and load transfers at greater depth. On the contrary, a
long free-head pile fails due to the insufficient moment capacity of the section, as the
large embedment depth of the pile does not allow the lower part to move or rotate. In
the case of a short fixed-head (capped) pile, the pile may fail by translation whereas
the failure occurs at cap level for the long fixed-head piles due to the development of
high bending stress. These observations ultimately lead to the development of
appropriate design methods by considering the short or long pile type with or without
the head restraint.

This section will first demonstrate how a pile can be classified as a short or long pile.
Soil-pile interaction is considered using stiffness factors (R for OC clay and T for
sands, NC clay) for this classification purpose.

For soils with constant modulus with depth (i.e., stiff OC clay), the stiffness factor “R”
is defined as:

𝑅 (in units of length)

Where, EI=pile stiffness; k≈(k1/5) and k1=modulus of horizontal subgrade reaction

using 305mm square plate (Terzaghi’s values are given below), B=width of the pile.

Consistency Firm to stiff Stiff to very stiff Hard

Undrained shear 50-100 100-200 >200

strength (cu), kPa

Range of k1 15-30 30-60 >60

(MN/m3)

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C3
Topic 3C: Laterally loaded piles

Soil modulus, 3-6 6-12 >12

k=k1/5 (MN/m3)

For soils with linear increase of modulus (i.e., soft NC, sands) the stiffness factor “T”
is defined as:

𝑇 (in units of length)

Where, k=nh.x/B and nh=coefficient of modulus variation, x=depth

nh for various soils:

• Soft NC: 350-700 kN/m3

• Soft organic silts: 150
kN/m3
For sands: Terzaghi is found
to be conservative (Reese et
al, 1974)

Effect of saturation on nh is to
be considered.

The above stiffness factors can be used to classify the free-head piles in a constant
or linearly increasing soil modulus case as per the table below (Tomlinson &
Woodward, 2008).

Pile type Soil modulus

Linearly increasing Constant

Rigid (free-head) L ≤ 2T L ≤ 2R

Elastic (free-head) L ≥ 4T L ≥ 3.5R

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C4
Topic 3C: Laterally loaded piles

The lateral load calculation methods can be classified in to two broad categories:

(i) Methods to determine the ultimate lateral load capacity

(ii) Methods to determine lateral load corresponding to an allowable deflection

 Brinch Hansen’s method (based on earth pressure): Only applicable to short

piles in c- and layered soils.
 Broms Method (based on earth pressure): Applicable to both short and long
piles in only cohesive or cohesionless soils. Maximum moment, shear and
deflection at pile head can be determined.
 Elastic method (based on subgrade reaction): Applicable to both short and long
piles in only cohesive or cohesionless soils. Distribution of moment, shear and
deflection along the length of pile can be determined.
 p-y method (based on variable stiffness, considers soil plasticity): Applicable to
short or long piles in cohesive, cohesionless or layered soils. Variations of
moment, shear force and deflection along length of pile can be determined.

2.1 Brinch Hansen’s method

Brinch Hansen’s (1961) method is suitable for short piles in uniform and layered soils.
It can be used to calculate the ultimate lateral resistance of the pile based on lateral
earth pressure theory and generally simple. Also, the method can be used for longer
semi-rigid piles to obtain a first approximation of the stiffness and embedment length
for a more rigorous analysis.

The resistance of the rigid pile is taken as the algebraic sum of the passive resistances
above and below the point of rotation “X” in the figure below (Tomlinson & Woodward,
2008). The location of the point of rotation is determined by trial-and-error solution.
For the calculation of passive resistance, the pile is divided into “n” number of
horizontal elements of length (L/n), where L is the length of the pile.

The unit passive resistance at a depth “z” below the ground surface can be calculated
using the following equation:

𝑝 𝑝 𝐾 𝑐𝐾
where, poz=effective overburden pressure, c=cohesion, Kqz &
Kcz=passive pressure coefficients for frictional and cohesive components at depth “z”.

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C5
Topic 3C: Laterally loaded piles

(a) Free-head (b) Fixed-head (c) SF/BM

By knowing z, B (width in the direction of rotation) and , the coefficients Kq and Kc

can be found using the figure below.

To find depth of the point of rotation “X”, take M=0 at the point of horizontal load
application.

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C6
Topic 3C: Laterally loaded piles

𝐿 𝐿
𝑀 𝑝 𝑒 𝑧 𝐵 𝑝 𝑒 𝑧 𝐵 0
𝑛 𝑛

By the process of trial and error, determine the depth “z” at which M=0.

If the pile head carries a moment then it can be replaced by a force and eccentricity
(M=H.e).

For a fixed-head pile the equivalent height (e1) of a force H acting on free head is
0.5(e+zf) where, zf=depth from ground surface to point of virtual fixity. Tomlinson &
Woodward (2008) provided the following guideline for the depth of point of virtual fixity
(zf).

For practical design purposes:

 Compact granular soil: zf=1.5m

 Stiff clay: zf=1.5m (below the depth of zone of shrinkage)
 Soft clay or silt: zf=3m
ACI recommends:

 Stiff OC clays (constant modulus): zf=1.4R

 NC clays, silt, granular soils & peat: zf=1.8T
Finally, the ultimate lateral load capacity (Hu) is obtained by taking moment about the
point of rotation “X”. The allowable lateral load capacity (Hall) of the pile can be
calculated as Hu/FoS. Subsequently, the bending moment and shear force variations
along the length of the pile are to be determined. The maximum bending moment
should not exceed the ultimate moment capacity of the pile.

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C7
Topic 3C: Laterally loaded piles

Example 1

An example taken from Tomlinson & Woodward Hu=??

(2008) is shown here for a deeper e=4m
understanding of Brinch Hansen’s method.

Find (i) Ultimate lateral load capacity (Hu) and

(ii) allowable load (Hall) for displacement
<25mm. L=6m
Solution:

Step 1: Find R as below. D=0.9m

.
𝑅 = 3.37m
. .
Ep=26GPa
For cu=120kPa (short-term)
k=(k1/5)=(36/5)=7.2MPa/m (From Table above)
and Ep=26GPa

Check L (6m)< 2R (2*3.37=6. 74m)

Stiff OC clay
cu(av)=120kPa;
Therefore, design the pile as a short pile.
o
c’=10kPa, ’=25
(i) Ultimate lateral load capacity (Hu)
Short term-undrained capacity:

 Piles are divided n=6 segments ie. L/n=6/6=1m and 1m width of pile
 For cu=120kPa and u=0 (short-term), find Kc values from figure at “z”
 The passive pressure variations with depth are tabulated below and the
distributions are shown in figure.

z (m) 0 1 2 3 4 5 6
z/B 0 1.1 2.2 3.3 4.4 5.5 6.6
Kc 2.2 5.5 6.2 6.7 7 7.2 7.3
cuKc 264 660 744 804 840 864 876

Try x=4m (depth of point of rotation “X”)

∑MH=(462x1x4.5)+(702x1x5.5)+(774x1x6.5)+(822x1x7.5)-[(852x1x8.5)+(870x1x9.5)]
= +1629 kN-m/m width of pile
Department of Civil Engineering, Monash University
Created: 6/2020
CIV4249: Foundation Engineering 3C8
Topic 3C: Laterally loaded piles

Try x=3.9m

∑MH=+297kN-m/m width of pile which can be considered as approaching zero value.

To find out Hu, take moment around the point of rotation “X” as:

∑MX=Hux7.9=(462x3.4)+(702x2.4)+(774x1.4)+(820.2x0.45)+(838.2x0.1x0.05)+(852x
0.6)+(870x1x1.6).

So, Hu=828 kN/m width of pile

For D=0.9m, Hu=0.9x828=745kN

Distributions of passive resistances and forces with depth (a) undrained-short term (b)
drained-long term (Tomlinson & Woodward, 2008)

Long term-drained capacity:

 Use c’=10kPa, ’=25o (long term)

 Find Kc, Kq from figure (values tabulated below)
 Find passive resistances with depth as below and distribution with depth is
shown in above figure.

z (m) 0 1 2 3 4 5 6
z/B 0 1.1 2.2 3.3 4.4 5.5 6.6
Kc 5.8 16 20 23 26 27 28

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C9
Topic 3C: Laterally loaded piles

c’Kc 58 160 200 230 260 270 280

Kq 3.3 5 5.5 5.9 6.3 6.7 6.9
po (kPa) 0 18.6 39.3 58.8 78.5 98.2 118
poKq 0 93 216 347 495 658 814
(kPa)
c’Kc+poKq 58 253 416 577 755 928 1094
(kPa)

By algebraically summing the moment around the load application point (similar to the
undrained calculation) and varying the x-value,

 x=4.5m gives moment zero about H

Now take moment about “X” (rotation point) to find Hu=530 kN/m width of pile

 For D=0.9m, Hu=0.9x530=477kN

(ii) Lateral deflection (Structural consideration)

A simple method of estimating lateral deflection of piles is:

 Assume pile is fixed at an arbitrary depth below ground surface (zf values
were given previously)
 Then calculate deflection as for simple cantilever with a free or fixed head

Pile head deflection for a free-headed pile (y)= :

Pile head deflection for fixed headed pile (y)=

Allowable load (Hall) based on deflection:

 Limit of lateral deflection at ground level=25mm

 Depth of point of fixity=zf=1.4R (stiff OC clay)
 Pile head deflection (free head) equation can be used with R=3.37m,
zf=1.4*3.37=4.72m, E=26GPa, I=D4/64, e=0 (at ground level)
 H=0.025*3*26x106*(x0.94/64)/(4.72)3=597kN > 477kN
 Soil resistance governs the capacity

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C10
Topic 3C: Laterally loaded piles

Allowable lateral capacity of the pile (Hall)=Hu/FoS=477/1.5=318kN

Note: The ultimate lateral capacity (Hu) for a long pile depends on the pile bending
capacity rather than soil yielding due to the very large passive resistance. This is a
structural design and one has to follow the structural design guidelines (perhaps
discussed in CIV2226) for determining the ultimate lateral load capacity by knowing
the ultimate moment capacity (Mu) as follows.

𝐻 [for free-headed long pile]

𝐻 [for fixed-headed long pile]

2.2 Broms method

Broms (1965) introduced his method for calculating the ultimate lateral load of short
(rigid) piles based on simplified earth pressure distributions along the length of the
pile. The ultimate capacity can be calculated for a single layer cohesive or frictional
soil but cannot be used for c- soils or multi-layered soils. The pressure distribution
along the length of the in cohesive and frictional soils can be found in standard texts
and Das & Sivakugan (2019).

By knowing the pile head conditions (free or fixed), short (rigid) or long (flexible) piles,
eccentricity, and type of soils (clay or sand), the ultimate lateral load capacity can be
determined using the design charts provided below (taken from Das & Sivakugan,
2019). The deflection of the pile head xz (z=0) under working load can be estimated
using Broms (1965) charts available in standard texts (Das & Sivakugan, 2019).

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C11
Topic 3C: Laterally loaded piles

(a) In sand (b) In clay

Design chart for short piles in (i) sand and (ii) clay (Das & Sivakugan, 2019).

(a) In sand (b) In clay

Design chart for long piles in (i) sand and (ii) clay (Das & Sivakugan, 2019).

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C12
Topic 3C: Laterally loaded piles

2.3 Elastic method

In general, the allowable lateral load capacity of a pile depends on the acceptable pile
head deflection. Matlock & Reese (1960) proposed a method based on the subgrade
reaction, where the simpler Winkler soil model was considered.

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C13
Topic 3C: Laterally loaded piles

The values of different coefficients used in the above solutions can be found in the in
Das & Sivakugan (2019). Readers are advised to refer this text for further details.

Elastic solutions for sands (fixed head pile)

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C14
Topic 3C: Laterally loaded piles

2.4 p-y method for lateral capacity

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C15
Topic 3C: Laterally loaded piles

p-y method step by step:

 Step 1: Establish a set of p-y curves for the soil-pile system (separate
procedure for sand and clay, follow standard texts)
 Step 2: Calculate characteristic lengths T (linearly increasing modulus, sands)
or R (constant modulus, stiff clay) for a given or estimated nh or k.
 Step 3: Determine deflection with depth based on T or R using subgrade
reaction methods
 Step 4: Determine p with depth from the p-y curves based on the calculated
deflections
 Step 5: Find k=p/y then nh=k/x or k1=k and finally T=(EI/nh)0.2 or R=(EI/k)0.25.
 Step 6: Check T or R value with Step 2. If they match then fine, otherwise go
for second trial with k or nh from Step 5. Continue until assumed and
calculated values agree. Use final R or T for y and M.
Establishment of p-y curves (sands)

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C16
Topic 3C: Laterally loaded piles

Establishment of p-y curves (clay)

Example 2 (Broms & p-y method)

Lateral pile in sand

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C17
Topic 3C: Laterally loaded piles

Department of Civil Engineering, Monash University

Created: 6/2020
 CIV4249: Foundation Engineering 3C18
Topic 3C: Laterally loaded piles

Example 3

p-y method for submerged clay

Laterally loaded piles in submerged soft clay e=2m P= 400 kN

EI= 435000 kn-m^2
Table Representatiive values of  50 (Table 9.7, Murthy-Adv. Foundation Engineering)
cu= 20 kPa d (m)= 0.61 m 2m p-y curve for soft clay c (kPa)*  50 (%)
eff. Unit wt.= 8.75 kN/m^3  50 (table)= 0.02 Soft clay under water 12-25 2
1.0
Depth, x= 0.5 m p/pu 25-50 1
pult(1)= 44 kN/m Work space for calculating pult p/pu =0.5(y/y50)^0.33 50-100 0.7
pult(2)= 110 kN/m pu1=(3+ˠ'x/c+0.5x/d)cd 100-200 0.5
pul t at x, min(pu 1,2)= 44 kN/m pu2=9cd 200-400 0.4
y50= 0.0305 m y50=2.5d 50 8 y/y50 *: range adjusted slightly

pul t constant at 8y50= 0.244 m

pu at different depths to be used for p-y curves constructions
120
pu at 0.5m= 44 kN/m
>=6m
pu at 1m= 52 kN/m 20m
100 4m
pu at 2m= 67 kN/m
Pu at 4m= 98 kN/m
pu at 6m= 110 kN/m 80

p (kN/m)
pu at 8m= 110 kN/m L= 20 m 2m
60
1m
0.5m
Construction of p-y curves p/pu=0.5(y/y50)^0.33 use this eqn to find p/pu ratio relevant to y/y50 40
y/y50 y(m) p/pu p(2m) p(4m) p(6m) p(1m) p(0.5m)
0 0.000 0.00 0 0 0 0 0 20
0.25 0.008 0.32 21 31 35 16 14
0.5 0.015 0.40 27 39 44 21 18 8y50
Select 0
Es, kPa
1 0.031 0.50 34 49 55 26 22 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350
intervals 0 500 1000 1500 200
2 0.061 0.63 42 62 69 33 28 y (m) 0
4 0.122 0.79 53 77 87 41 35 trial 1
1
8 0.244 0.99 67 97 109 52 44 trial 2
10 0.305 0.99 67 97 109 52 44 2
1st TRIAL, assume nh= 500 kN/m^3 T= 3.9 m y=Ay(PT^3/EI)+By(MT^2)/EI General eng for deflection 3

Depth (m)
Range (nh) for soft NC clay=278-556 kN/m^3 Zmax=L/T 5.2 >4 Long pile 4
(Davisson & Prakash, 1963) y=yA+yB=(PT^3/EI)Ay+(MT^2/EI)By
5
TRIAL 1 From Table (refer books)-Reese & Matlock p-y curve kN/m/m
depth, m Z=L/T Ay By yA (mm) yB (mm) y=yA+yB depth, m y,m p (kN/m) Es=p/y 6
0 0 2.435 1.623 130 45 175 0 0.17 0 0 7
0.5 0.13 2.21 1.37 118 38 156 0.5 0.16 38 244 8 y = 0.0044x
1 0.26 2.03 1.215 108 33 142 1 0.14 41 289 y = 0.0054x
9
2 0.52 1.644 0.874 88 24 112 2 0.11 51 456
4 1.03 0.962 0.364 51 10 61 4 0.06 60 978 kN/m^3 m
6 1.55 0.46 0.81 25 22 47 6 0.05 62 1323 Trial nh T
8 2.07 0.142 -0.07 8 -2 6 8 0.01 10 1770 1 500 3.9
10 2.58 0.0335 -0.0795 2 -2 0 10 0.00 2 227 4.5
T= 4.5 m 3 200 4.7 close to 4.5
TRIAL 2 use nh = 227 kN/m^3 p-y curve kN/m/m
depth, m Z=L/T Ay By yA (mm) yB (mm) y=yA+yB depth, m y,m p (kn/m) Es=p/y
0 0 2.435 1.623 209 61 270 0 0.27 0 0
y (m) and M (kN-m)
0.5 0.11 2.26 1.43 194 54 248 0.5 0.25 44 178
0 500 1000 1500 2000 2500
1 0.22 2.07 1.263 177 48 225 1 0.23 50 222 0
2 0.44 1.7 0.935 146 35 181 2 0.18 59 326
4 0.88 1.08 0.44 93 17 109 4 0.11 75 687
6 1.32 0.64 0.16 55 6 61 6 0.06 65 1067
8 1.76 0.24 -0.03 21 -1 19 8 0.02 30 1543 Deflection (y)
10 2.21 0.0335 -0.0795 3 -3 0 10 0.00 5 Moment (M)

TRIAL FINAL use nh = 200 kN/m^3 T= 4.7 m kN-m p-y curve kN/m/m
depth, m Z=L/T Ay By yA (mm) yB (mm) y=yA+yB Am Bm M depth, m y,m p (kn/m) Es=p/y
Depth, m

0 0 2.435 1.623 225 65 290 0 1 800 0 0.29 0 0

10
0.5 0.11 2.26 1.43 209 57 266 0.1 1 986 0.5 0.27 44 165 P=400 kN
1 0.22 2.07 1.263 191 50 242 0.2 0.997 1170 1 0.24 52 215
2 0.43 1.7 0.935 157 37 194 0.419 0.982 1565 2 0.19 61 314 2m GL
4 0.86 1.08 0.44 100 17 117 0.671 0.9 1968 4 0.12 78 665
6 1.29 0.64 0.16 59 6 66 0.767 0.71 1995 6 0.07 72 1098 15
8 1.72 0.31 -0.005 29 0 28 0.71 0.54 1753 8 0.03 58 2037
10 2.15 0.132 -0.067 12 -3 10 0.567 0.352 1336 10 0.01 20 2096
15 3.23 0.0216 0.0556 85
20 4.30 -0.01 0.0388 12
20 20m

(DRAFT release) Please check the formulas and output prior confirmation of the whole calculation.

3. Laterally loaded piles in a group

Piles behave as single unit if they are:

 Spaced at > 3B in direction normal to direction of loading, and

 > 6-8B direction parallel to the direction of loading
If less spacing then check capacity for block using width of the group

Department of Civil Engineering, Monash University

Created: 6/2020
CIV4249: Foundation Engineering 3C19
Topic 3C: Laterally loaded piles

References

Das, B.M. & Sivakugan, N. (2019). Principles of foundation engineering. Cingage,

USA.

Tomlinson, M. & Woodward, J. (2008). Pile design & construction practice. Taylor &
Francis, NY, USA.

Prakash, S. & Sharma, H.D (1990). Pile foundations in engineering practice. John
Wiley & Sons Inc.

Department of Civil Engineering, Monash University

Created: 6/2020