Sie sind auf Seite 1von 22

Atkinson, J. H. (2000). Geotechnique 50, No.

5, 487508

Non-linear soil stiffness in routine design


J. H . AT K I N S O N 

Soil stress-strain behaviour is highly non-linear and this has Le comportement contrainte-deformation du sol est extreme-
an important inuence on the selection of design parameters ment non lineaire et ce fait inuence considerablement le choix
for simple routine geotechnical calculations. Non-linear be- des parametres de design pour les simples calculs geotechni-
haviour can be characterized by rigidity and degree of non- ques de routine. Le comportement non lineaire peut etre
linearity and these can be determined from measurements of caracterise par la rigidite et le degre de non linearite, deux
very small strain-stiffness, peak strength and failure strain. parametres qui peuvent etre determines en mesurant une tres
Very small strain-stiffness can be found from measurements faible deformation-rigidite, la resistance maximale et la defor-
of shear wave velocity in situ or in laboratory tests. Peak mation a la rupture. On trouve une tres faible deformation-
strength and failure strain can be measured in routine rigidite en mesurant la velocite de l'onde de cisaillement in situ
laboratory tests but are strongly inuenced by initiation of ou dans les essais en laboratoire. La resistance maximale et la
shear bands. The important non-linear stiffness parameters deformation a la rupture peuvent etre mesurees par des essais
for soil are related to its composition and to its current de routine en laboratoire mais sont fortement inuencees par
state. Back-analyses of the load settlement behaviour of full- l'initiation de bandes de cisaillement. Les parametres impor-
scale and model foundations demonstrate the inuence of tants de rigidite non lineaire pour le sol sont lies a sa composi-
non-linear soil behaviour. Variations of stiffness with settle- tion et a son etat actuel. Des retro-analyses du comportement
ment calculated from full-scale and model foundations agree de tassement pour des fondations grandeur nature et des
well with non-linear soil stiffnesses based on rigidity and maquettes montrent l'inuence du comportement non lineaire
degree of non-linearity. These results suggest a simple meth- d'un sol. Les variations de rigidite en fonction du tassement,
od for routine design which takes account of soil non-linear calculees d'apres des fondations reelles et des maquettes,
stiffness. correspondent bien a la rigidite non lineaire du sol basee sur
la raideur et le degre de non linearite. Ces resultats suggerent
KEYWORDS: Design footings/foundations; in situ testing; labora- une methode simple pour les etudes de routine, methode qui
tory tests; settlement stiffness tient compte de la rigidite non lineaire du sol.

INTRODUCTION strengths depending mainly on drainage and strain and the peak
In the Autumn of 1969 when I started research at Imperial strength is appropriate for characterizing non-linearity. Peak
College on soil stiffness I had three textbooks. These were: Soil strengths are associated with slip planes or shear bands and it is
mechanics in engineering practice (Terzaghi & Peck, 1948), necessary to consider the inuence of these on strength meas-
The measurement of soil properties in the triaxial test (Bishop ured in laboratory tests.
& Henkel, 1957) and Critical state soil mechanics (Schoeld &
Wroth, 1968). As a young research student it was difcult to
understand that these three books were all dealing with soils in Non-linear behaviour of soil
ground engineering. Two important themes of my work have One of the major problems in ground engineering in the
been to try to clarify the principal issues covered in these three 1970s and earlier was the apparent difference between the
books and to research soil strength and stiffness. stiffness of soils measured in laboratory tests and those back-
It is now well known that the stressstrain behaviour of soil calculated from observations of ground movements (e.g. Cole &
is highly non-linear and soil stiffness may decay with strain by Burland, 1972; St John, 1975; Wroth, 1975; Burland, 1979).
orders of magnitude. This means that for a geotechnical struc- These differences have now largely been reconciled through the
ture such as a foundation, retaining wall or tunnel, soil stiffness understanding of the principal features of soil stiffness and, in
varies both with position and with loading. particular, the very important inuence of non-linearity. This is
Many aspects of non-linear soil stiffness are now well under- one of the major achievements of geotechnical engineering
stood. They have been incorporated into numerical models and research over the past 30 years.
have been used with success in geotechnical design. Many of Figure 1 illustrates a typical stiffness-strain curve for soil. At
these non-linear models and numerical analyses are relatively small strains the stiffness is relatively large; at strains close to
complex and require special testing and lengthy calculation. failure the stiffness is small: this is soil being non-linear. Fig. 1
There are, however, many practical cases for which these includes typical ranges of strain for laboratory testing and for
complex models and analyses are not justied and familiar structures. The ranges of strain for the different testing techni-
methods based on load factors or simple elastic analyses are ques in Fig. 1 are similar to those given by Atkinson & Sallfors
sufcient. These may be improved if allowance is made for soil (1991). These will be discussed later in more detail. The typical
non-linearity. strain ranges for structures are those given by Mair (1993). A
The principal purposes of the 40th Rankine Lecture, and of typical characteristic strain in the ground is 01%; this repre-
this paper, are to consider how soil non-linearity can be sents a movement of 10 mm across a gauge length of 10 m.
quantied from the results of relatively simple tests and to Generally, strains in the ground will vary from zero far away
examine the inuences of soil non-linearity on simple routine from the structure to relatively large values near the structure
design methods. In characterizing non-linearity it is necessary and at the edge of a rigid foundation they will be very large.
to consider both stiffness, strength and strain at failure and the The typical strain ranges proposed by Mair (1993) were based
relationships between them. Measurement of soil stiffness over on stiffnesses which gave reasonable designs for structures in
the full range of loading from very small strain to failure London Clay.
requires the use of local strain gauges but stiffness at very small
strain can be determined relatively easily from measurements of
shear wave velocity in laboratory tests or in situ. Soil has many Routine design
In geotechnical engineering there are some works which
 Professor of Soil Mechanics, City University, London. require detailed analysis either because there are special design
487
488 ATKINSON

Typical strain ranges


Retaining walls
Stiffness: G

Foundations

Tunnels

00001 0001 001 01 0 10


Shear strain, s: %
Dynamic methods

Local gauges

Conventional soil testing

Fig. 1. Characteristic stiffnessstrain behaviour of soil with typical Fig. 2. Methods for routine design of simple foundations
strain ranges for laboratory tests and structures (after Atkinson &
Sallfors, 1991 and Mair 1993)
where r is the change of settlement due to a change of
bearing pressure, , B is the width of the foundation, is
requirements or because there are substantial economies to be Poisson's ratio, I r is an inuence factor which depends princi-
made. An example would be the design of a large retaining wall pally on the geometry of the foundation (Poulos & Davis, 1974)
in an urban environment. In this case, it would probably be and Es is the secant Young's modulus corresponding to the
necessary to calculate the distribution of horizontal and vertical increment of loading. Es may be related to Young's modulus for
ground movements in front of and behind the wall, stresses in very small strain Eo through a stiffness ratio Es =Eo . Again,
the wall and loads in anchors or props both during construction partial factors may be applied to account for uncertainties.
and in service. These simple routine methods may also be applied to the design
Detailed analysis and design of a major geotechnical struc- of deep foundations, retaining walls and tunnels.
ture will require special laboratory testing involving the applica- Load factor Lf and stiffness ratio Es =Eo are design para-
tion of complex stress paths and the measurement of small meters. They will depend on, among other things, the soil, its
strains together with numerical analyses using soil models state and its stressstrain behaviour, the structure and the design
which take account of the important features of soil behaviour, movements. Since these simple routine design methods aim to
including current state, recent history, in-elastic deformations, determine only one movement it must always be possible to
anisotropy, general stress states, rotation of axes of stress and select load factors or stiffness ratios which give correct solu-
strain, and so on (Hight & Higgins, 1995). All this is very tions. If the soil is non-linear then these will vary with loading
complicated and demanding and requires special equipment and and movement or strain.
expertise to obtain reliable solutions. It must be emphasized that these simple methods are, of
There are, however, very many cases where it is not so course, limited. They can work only for calculating one move-
important to have such detailed analyses and where relatively ment in one direction for relatively simple structures and well-
simple solutions are all that are needed. These routine analyses behaved soils. If more information is required, such as both
calculate only one movement in one direction; examples would vertical and horizontal movements or a prole of settlement or
be the settlement of a foundation, the horizontal movement at distributions of stress in the ground, then much more compli-
the top of a simple retaining wall, the surface settlement above cated analyses will be required.
the centre-line of a tunnel and so on. The simple methods described here are applicable to drained
Figure 2 illustrates the settlement of a loaded shallow or to undrained loading, making use of data from drained or
foundation and the two principal methods for routine design. undrained tests but not to cyclic or repeated loadings. They are
The general principles apply also to the design and analysis of applicable to soils which have the characteristic behaviour
simple retaining walls and tunnels. For the shallow foundation described later. These include relatively stiff ne and coarse
illustrated the basic requirement is to determine the design grained soils which are not strongly bonded; they exclude very
bearing pressure d which will cause a design settlement rd . soft soils, strongly bonded soils, and soft rocks and soils with
In the rst method the allowable bearing pressure a is unstable structure.
calculated from
1
a Lf c c (1) CHARACTERISTICS OF NON-LINEARITY
Fs Figure 3 denes basic strength and stiffness parameters for a
where c is a calculated ultimate bearing capacity, Lf is a load triaxial test. In Fig. 3(a) the cylindrical sample has axial and
factor and Fs is a factor of safety where Fs 1=Lf . In this radial stresses a and r and strains a and r and the deviator
method the factor of safety or load factor is there to limit stress is q ( a r ). On loading there is a non-linear stress-
settlements; the intention is to reduce the ultimate bearing strain curve as illustrated in Fig. 3(b). The sample fails at the
capacity by a factor so that the design point is in the part of the peak deviator stress qf at a strain f . The stiffness is Young's
load settlement curve where settlements are relatively small. modulus E which may be dened as a tangent Et or as a secant
Additional partial factors may be applied to various actions and Es . The stiffness at very small strains near the start of loading
reactions. is Eo .
In the second method the settlement is calculated from The tangent and secant Young's moduli vary with strain as
illustrated in Fig. 3(c). There are three regions dened by
Atkinson & Sallfors (1991). In the very small strain region the
r (1 2 ) stiffness is approximately constant and Et Es Eo and this
Ir (2)
B Es region is limited by a strain o . There is a small strain region
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 489
q q

qf
qf

E Eo
f r f

E E

Eo
Eo

f f
r f = 2r
(a) (b)

Fig. 4. Stressstrain behaviour of simple materials: (a) linear


material; (b) non-linear material
Fig. 3. Simple stiffness parameters for non-linear soil

is not the strain at any characteristic point during the loading.


from o up to a strain of about 01% within which the stiffness Non-linear materials fail at strains f which are greater than r
decays rapidly. There is a large strain region beyond 01% and the ratio f =r nl is a measure of the degree of non-
within which the stiffness is relatively small. The strain linearity. For the particular case in which the stiffness decreases
0:1% coincides with characteristic strains in the ground linearly with strain nl 2 because the area beneath the stiff-
near structures and with the smallest strain that can be meas- ness-strain curve must equal the strength as shown in Fig. 4(b).
ured reliably in conventional soil tests. The secant modulus Es Rigidity and the degree of non-linearity together serve to
continues to decrease gradually but remains positive even characterize non-linear stressstrain behaviour. The parameters,
beyond the peak deviator stress while the tangent modulus Et is very small strain stiffness Eo , strength qf , and failure strain f ,
zero at failure and then becomes negative as the soil softens. If are all easily measured.
the stressstrain axes are q and s ( 2=3(a r )) the stiffness Table 1 summarizes the stiffnesses and strengths of some
dq=ds 3G where G is the shear modulus. common materials in order to put the properties of typical soils
Figure 4(a) illustrates the stressstrain behaviour of a simple into context. The values are approximate and, for brittle materi-
material which has a linear stressstrain response and which als, they are for compression. The values given for soft and stiff
fails at a deviator stress qf with the corresponding stiffness- soils were obtained from test results and simple correlations, as
strain curve. The rigidity is dened as the ratio of stiffness to described later. There are some interesting values in Table 1.
strength, E=qf , and this is equal to 1=f . The ratio of stiffness The rigidity of soil is greater than that of other common
to strength is an important parameter. It appears in solutions for materials largely because soil is relatively very weak. The
cavity expansion (Vesic, 1972). For linear materials it deter- rigidity of stiff soil is less than that of soft soil. This is a
mines the failure strain and characterizes brittleness or ductility. surprising result which will be discussed later. The degree of
Vesic (1972) dened the ratio of shear modulus to undrained non-linearity of soil is highly variable and covers almost the
strength G=su as rigidity index. whole range of all the other materials. It is this variation in
The area beneath the stiffness-strain curve is Ef and this is degree of non-linearity which characterizes soil stiffness and
equal to the strength qf . Simpson (1992) showed that this is a which makes geotechnical design demanding.
general result and holds for non-linear materials and for drained
and undrained loading.
Figure 4(b) illustrates the stressstrain behaviour of simple MEASUREMENT OF STRAIN AND STIFFNESS IN LABORATORY
non-linear materials. The rigidity is now dened as Eo =qf TESTS
1=r where r is a reference strain: it is important to note that The values of strength, stiffness and permeability parameters
the reference strain r is simply dened from the rigidity and it measured in laboratory tests depend on many factors including:

Table 1. Properties of some common materials


Material Eo : MPa qf : MPa Eo =qf r : % f : % f =r
Concrete 28 000 40 700 015 035 2
Glass 70 000 1000 70 15 15 1
Mild steel 210 000 430 500 02 30 150
Copper 120 000 200 600 015 35 250
Aluminium 70 000 100 700 015 10 70
Rubber 10 20 05 200 800 4
Timber 10 000 20 500 02 5 25
Soft soil 100 005 2000 005 10 200
Stiff soil 300 03 1000 01 1 10
490 ATKINSON
the quality of the sample, the procedures used to set up the
sample in the apparatus prior to testing, details of the loading
path and rate of testing, details of the design and performance
of the apparatus and instruments, procedures for analysis and
interpretation of the raw test data. Many of these factors were
considered by Hight (1998).
Baldi et al. (1988) found that major sources of error in the
measurement of strain and stiffness in triaxial tests, particularly
at small strains, were in bedding and seating errors and in
misalignments of the loading ram or load cell with the top
platen. These errors are avoided by the use of local gauges
attached to the sample. There is, however, still a lower limit of
strain which can be measured reliably using local gauges but
stiffness at very small strain can be measured using dynamic
methods.
Figure 5 shows the characteristic stiffnessstrain curve for
soil with the three regions dened by Atkinson & Sallfors
(1991) and shows the different laboratory equipment and test
procedures best able to measure stiffness in each region.
At strains in excess of about 01% secant stiffness can be
measured with reasonable accuracy in triaxial tests using dis-
placement gauges mounted in the conventional manner outside
the cell. The accuracy of measurement can be improved by
reducing the bedding, seating and misalignment errors
(Atkinson & Evans, 1985).
Reliable measurement of soil stiffness throughout the small
strain-region from strains of about 0001% up to about 01%
can really only be made using local gauges attached directly to
the sample (Jardine et al., 1984). If measurements of strain are
made outside the cell, the corrections to account for the errors
are often greater than the strains being measured. Local gauges
must operate satisfactorily in water or oil under pressure and
must remain stable and accurate for long periods. (Tests on ne
grained soils to measure stiffness, taking account of recent
history and other effects, often last several weeks.) There are a
number of recent state-of-the-art reviews on the measurement of
soil stiffness using local gauges (e.g. Scholey et al., 1995).
Figure 6 shows the original Imperial College hydraulic stress
path apparatus (Atkinson, 1973). There are displacement gauges Fig. 6. Imperial College stress path apparatus (Atkinson, 1973)
inside the cell. They are not strictly local gauges because they
are attached to the platens and not directly to the sample. This
apparatus represents the early days of research into soil stiffness
in the small strain range.
Figure 7 shows commercially available miniature displace-
ment transducers, measuring axial and radial strains in a sample
in an hydraulic stress path triaxial cell (Cuccovillo & Coop,
1997). These instruments are relatively simple and reliable and
they are capable of very high resolution. They can resolve
stiffnesses at strains of about 106 (i.e. 00001%) which, for
many soils, extends into the very small strain region. This
represents a resolution of 01 micron over a gauge length of
100 mm. It is hardly necessary to have more precise measure-
ments and the problems are mainly in mounting the gauges on
the sample and in interpreting the data.
Figure 8 shows an example of a modern hydraulic stress path
cell capable of applying a full range of triaxial stress paths with
axis and radial strains measured using miniature displacement

Fig. 5. Measurement of soil stiffness in laboratory tests Fig. 7. LVDTs used as local gauges
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 491
The shear modulus G of a material is related to the velocity
Vs of a shear wave through it by

G rV 2s V 2s (3)
g
where r is the density, is the unit weight and g is the
acceleration of the Earth's gravity. The strains generated by the
passage of a shear wave will be very small, generally less than
0001% (Dyvik & Madshus, 1985), and so the shear modulus
calculated from equation (3) will be Go , the stiffness at very
small strain. Since total and effective shear stresses are equal
Go G9o Guo .
For an isotropic elastic soil the effective stress elastic para-
meters are related by
E9o 2G9o (1 9) (4)
and, for undrained loading for which u 0:5
Euo 3Guo (5)
Fig. 8. PC-controlled stress path apparatus with local gauges
There are also indirect methods for measuring shear modulus
in dynamic tests such as laboratory resonant column tests
(Richart et al., 1970) and in situ measurement of Rayleigh
transducers. Tests in equipment of this kind would be required surface wave velocities (Abbiss & Ashby, 1983). Direct meas-
to measure the whole of the stiffnessstrain behaviour over the urement of shear wave velocity using laboratory bender element
full range of strain needed for full numerical analyses of ground tests or in situ down-hole and cross-hole tests are relatively
movements using complex constitutive models. It is, however, simple to perform and interpret.
still difcult to use local gauges routinely and there are few
engineers with the expertise required to specify and supervise
testing and who are able to interpret the results. Measurement Measurement of shear wave velocity in laboratory tests using
of soil stiffness over the whole range of strain using local bender elements
gauges is unlikely to be routine, at least for some time to come. A bender element is a piece of piezo-ceramic plate which
bends if a voltage across it is changed or, if bent by an external
force, the voltage across it changes. Bender elements are usually
MEASUREMENT OF STIFFNESS AT VERY SMALL STRAIN USING set into the top and bottom platens of a triaxial or oedometer
DYNAMIC METHODS cell and penetrate about 3 mm into the sample. One element is
While it is difcult to measure soil stiffness in the small vibrated by changing the voltage across it, shear waves propa-
strain range using local gauges, it is much easier to measure gate through the sample and vibrate the other element. The
soil stiffness in the very small strain range using dynamic input and output voltages are continuously recorded and the
methods. Early research in soil dynamics (e.g. Hardin & travel time determined.
Drnevich, 1972) was associated with ground vibrations and Figure 10 shows bender elements set into the platens of a
stiffnesses measured using dynamic methods were found to be triaxial apparatus and Fig. 11 shows a hydraulic triaxial cell
considerably larger than those measured using conventional equipped with bender elements. The input and output voltages
triaxial tests with external displacement gauges. Georgiannou may be recorded on an oscilloscope or in a PC with an
et al. (1991) showed that stiffnesses measured at small strains oscilloscope card. This equipment has been used routinely at
in triaxial tests using local gauges were of the same order as City University to determine shear wave velocity and Go in
those measured in dynamic tests. Dynamic and static stiffnesses soils and soft rocks (Viggiani, 1992; Jovicic, 1997).
have now been reconciled by understanding soil non-linearity Bender elements (Shirley & Hampton, 1978; Dyvik & Mad-
and it is clear that it is the magnitude of the strain and not the shus, 1985) were originally developed to measure shear wave
strain rate which most inuences soil stiffness. velocities in soft soils. The equipment was modied and devel-
The basic principles and methods for determining soil stiff- oped for testing stiff soils and soft rocks by Viggiani (1992)
ness at very small strain from direct measurements of shear and by Jovicic (1997). Bender elements have been installed into
wave velocity in laboratory and in situ tests are illustrated in
Fig. 9. In laboratory tests shear waves are generated and
detected by bender elements (Shirley & Hampton, 1978). In situ
shear waves generated at the surface or below ground are
detected by instruments in boreholes or pushed in probes.

Bender elements
in triaxial or Down
oedometer hole
samples

Cross
hole
(a) (b)

Fig. 9. Measurement of Go from shear wave velocity: (a) laboratory


tests; (b) in situ tests Fig. 10. Bender elements in the platens of a triaxial cell
492 ATKINSON
shear wave propagating vertically with horizontal vibration, Vhv
is the velocity of a wave propagating horizontally with vertical
vibration and Vhh is the velocity of a wave propagating horizon-
tally with horizontal vibration.
In laboratory tests in which bender elements are installed in
the end platens, the different velocities can be measured on
samples which are installed with different orientations (Simpson
et al., 1996). Alternatively, using the equipment developed by
Pennington (1999), bender elements installed across a diameter
measure the velocities of shear waves which propagate horizon-
tally with vibrations which may be either horizontal or vertical.
Anisotropic shear moduli are related to shear wave velocities
by:

Govh V 2vh (6)
g

Gohh V 2hh (7)
g
Fig. 11. Equipment for bender element testing
and these are the same for total and effective stress. For a
cross-anisotropic material with a vertical axis of symmetry,
which is stiffer for shearing in a horizontal plane than for
separate mounts attached to the sides of the sample so the shear shearing in a vertical plane, Gohh . Govh . For a homogeneous
waves propagate across the diameter (Pennington et al., 1997). material Govh Gohv , as both correspond to the same mode of
This equipment is highly portable and has been used to measure shearing and so Vvh Vhv . In practice, values of Vvh and Vhv
shear wave velocities in unconned samples immediately after measured in situ are often different and this is commonly
recovery from the ground. attributed to horizontal layering (Simpson et al., 1996).
The effective distance travelled by the shear wave through The introduction of elastic anisotropy, even of the simplest
the sample is the tip to tip distance between the bender kind, complicates the analyses which relate the anisotropic
elements (Viggiani, 1992). The quality of the measurement of elastic shear moduli to the other elastic parameters. For a cross-
the travel time is sensitive to the form, frequency and amplitude anisotropic elastic material there are ve independent elastic
of the shear wave. In early tests, the excitation was normally a parameters. Although both Govh and Gohh can be obtained from
single square pulse that generated unwanted near-eld effects. measurements of shear wave velocities in situ or in laboratory
Nowadays the excitation is normally a sine wave and travel tests, determination of all ve effective stress elastic-parameters
times can be measured reliably (Jovicic et al., 1996). for a cross-anisotropic material requires additional measure-
In principle, the tests are sufciently simple to perform and ments (Lings et al., 2000).
the results are sufciently reliable for routine analyses. On a For undrained loading, the condition of constant volume
scale of cost and difculty, simple bender element tests to imposes restrictions on the values for the undrained anisotropic
determine shear wave velocity in a sample should fall between Poisson's ratios (Gibson, 1974). If the degree of anisotropy for
conventional unconsolidated undrained (total stress) triaxial tests undrained loading is N (Atkinson, 1975) then
and consolidated drained or undrained (effective stress) tests: Euv uvh
they are considerably less costly and easier to perform than N (8)
stress path tests using local gauges. Euh uhv

With zero volumetric strain and equal horizontal strains


Direct measurement of shear wave velocity in situ 1
Measurement of the velocities of waves propagating through uvh (9)
the ground is a well-established technique in geophysics and is 2
used in ground investigations mainly for proling. In common uhv uhh 1 (10)
applications the source generates both shear waves (S-waves)
and compression waves (P-waves) and it is the faster travelling and, from equation (8)
P-waves which arrive rst and mask the arrival of the S-waves. 1
In order to measure stiffness it is necessary to have a source uhv (11)
that generates S-waves which are essentially free from P-waves 2N
 
and other fast travelling waves (Jovicic et al., 1996). 1
uhh 1 (12)
For in situ measurement of shear wave velocity, shear waves 2N
are normally propagated vertically downwards from the surface
(down-hole tests) or horizontally from one bore hole to other From equation (4), with the appropriate undrained anisotropic
bore holes (cross-hole tests), as illustrated in Fig. 9(b). Alter- elastic parameters, and with equation (12)
natively, a seismic cone (Campanella et al., 1986) can be used  
1
in place of geophones in bore holes. The travel times are Euh 2Guhh (1 uhh ) 2Guhh 2 (13)
normally recorded between two receivers rather than from the 2N
source. It is important to determine the locations of the recei-
vers in order to obtain accurate measurements of the wave path and, from equation (8)
length. Euv NEuh (14)
Thus, the undrained anisotropic Young's moduli and Poisson's
ANISOTROPY IN SHEAR WAVE VELOCITY MEASUREMENTS ratios can be found from measurements of Vhh (giving Guhh )
Many soil properties are anisotropic and shear wave velocity together with a value for N.
depends both on the direction of propagation and on the The degree of anisotropy for undrained loading in terms of
direction of vibration. Horizontally, propagating shear waves total stress N is not the same as the degree of anisotropy for
may vibrate in a horizontal or in a vertical plane, as illustrated effective stresses n E9v =E9h (Atkinson, 1975) and neither is it
in Fig. 9(b), and these detect anisotropy. Vvh is the velocity of a the same as the ratio Govh =Gohh measured from shear wave
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 493
velocities. The measurement of N in the small and very small
strain ranges requires precise measurement of strains using local
gauges. If the degree of anisotropy for undrained loading N
remaines essentially constant with strain values for N can be
obtained from measurements of Euh and Euv at large strain in
conventional undrained triaxial tests on oriented samples.
If the soil is assumed to be isotropic, values for Young's
modulus for simple routine designs have to be calculated from
values of Go , obtained from measurements of shear wave
velocity using equations (4) or (5) with reasonable values of
Poisson's ratio. In this case, moduli should be determined that,
as far as possible, match the directions of signicant loading or
straining in the structure being designed. If the soil is assumed
to be cross-anisotropic, undrained total stress elastic-parameters
can be determined from equations (13) and (14) using values of
Gohh , calculated from measurements of Vhh using equation (7)
and values of N from conventional undrained triaxial tests. For
drained loading, determination of the effective stress cross- Fig. 13. Parameters for current state
anisotropic elastic-parameters requires additional measurements
(Lings et al., 2000).

state line, and Sv is equivalent to (Schoeld & Wroth,


VALUES FOR STIFFNESS OF SOIL AT VERY SMALL STRAIN 1968) and to the state parameter dened by Been & Jefferies
Figure 12 shows values for Go determined from shear wave (1986).
velocities measured in situ and in laboratory bender element If the reference line is a normal compression line, states
tests on undisturbed samples at the in situ stress state, for a below and to the left are overconsolidated. A ne grained soil
variety of soils ranging from soft silt and clay to stiff clay and can reach an overconsolidated state only by virgin compression
till. In all cases the in situ and laboratory values agree well. down the normal compression line followed by unloading and
The measured stiffnesses range between Go 10 MPa swelling. The state of a coarse grained soil can move directly
(Euo 30 MPa) to Go 120 MPa (Euo 360 MPa). from B to A by vibration or compaction at constant effective
stress and so compaction has the same inuence on state as
loading and unloading.
Soil state
In order to understand soil stiffness (and the whole of soil
mechanics) it is necessary to take account of the current state Variation of very small strain-stiffness with state for soils
of the soil. For isotropic stresses, the state is described by the The variation of very small strain-stiffness with stress, speci-
current effective stress and specic volume with respect to a c volume and overconsolidation has been investigated exten-
reference line; for anisotropic stress states, the stress ratio sively in the past. A number of relationships have been
should be included. Fig. 13 illustrates soil states with axes proposed (e.g. Hardin, 1978) and most take the general form:
specic volume (or water content) and effective stress. The Go Af (v) p9 n Rom (15)
reference line may be the projection of the critical state line or
one of the normal compression lines (Chandler, 2000). At a where f (v) is some function of specic volume, p9 is the
state at A the behaviour will be different to that at B at the current effective stress, Ro is the overconsolidation ratio dened
same effective stress and different to that at C at the same water as p9m = p9 where p9m is the maximum past effective stress. A, n
content. However, after normalization, by dividing by the cur- and m are material parameters. If the overconsolidation ratio Ro
rent stress, the states at B and C are equivalent. Consequently, is dened with respect to a normal compression line, the state
states along any line parallel with the reference line will be can be dened by only two of v, p9 and Ro and equation (15)
equivalent and so state is measured by the distance from the can be written as:
reference line.  n
Go p9
The state may be described by a stress state parameter S or A Rom (16)
a volume state parameter Sv. If the reference line is the critical pa pa

where pa is a reference pressure to make equation (16) dimen-


Go MPa measured in laboratory bender element tests

sionally consistent. ( pa , which inuences the value of A, is


140 normally taken as 1 kPa or as atmospheric pressure.)
Figure 14 shows results obtained from a set of bender
120
element tests on samples of reconstituted kaolin clay carried out
by Viggiani (1992). Fig. 14(a) shows results from tests on
100 Pentre silt normally consolidated samples for which Ro 1. The data fall
close to a line given by equation (16), with n 0:65 and
80 Bothkennar A 2000 with pa 1 kPa. Fig. 14(b) shows the results from
soft clay tests on overconsolidated samples with Go normalized by Gonc,
60 Chattenden; which is the value of Go for a normally consolidated sample at
London Clay the same effective stress. The data fall close to a line given by
40 Madingley; equation (16) with m 0:2.
Gault Clay
Figure 15 shows results obtained from a set of bender
Cowden Till
20 element tests on samples of a carbonate sand carried out by
Jovicic (1997). The behaviour of the carbonate sand is similar
0
0 20 40 60 80 100 120 to that of kaolin clay shown in Fig. 14, except the values of
Go MPa measured in situ the parameter m are different for samples which reached
states inside the normal compression line by loading and
Fig. 12. Go measured in situ and in laboratory tests (after Butcher, unloading (true overconsolidation) or by compaction before
2000) loading.
494 ATKINSON
2

105

Go/Gonc
Go: kPa

n
1

104
10 102 103 1 2 5 10
p : kPa Ro
(a) (b)

Fig. 14. Variation of Go with state for reconstituted kaolin clay: (a) normally consolidated samples; (b) overconsolidated
samples (Viggiani & Atkinson, 1995)

106
Go/Gonc
Go: kPa

105 Compacted

Truly overconsolidated

102 103 104 1 10 100


p : kPa Ro
(a) (b)

Fig. 15. Variation of Go with state for carbonate sand (Jovi i & Coop, 1997)

The carbonate sand had relatively weak grains and reached and ne grained soils. They also found that the values of the
states on a well-dened linear normal compression line at material parameters for coarse grained soils could be closely
stresses in excess of about 100 kPa above which considerable approximated by A 4000 and n 0:58, while the value of m
changes of grading were observed. Consequently, the gradings depended on the history of overconsolidation or compaction.
of truly overconsolidated samples differed from the gradings of It should be noted that, in order to determine values of
compacted samples which accounts for the different values for overconsolidation ratio Ro it is necessary to establish a true
the parameter m. normal compression line. For most coarse grained soils this will
require compression to very large effective stress (Coop & Lee,
1993).
Material parameters for very small strain stiffness
The parameters A, n and m in equation (16) are material
parameters and so they should depend on the nature of the STIFFNESS OF SOIL AT VERY SMALL STRAIN: SUMMARY
grains. Viggiani (1992) carried out bender element tests on The stiffness of soil at very small strains can be determined
reconstituted samples of a variety of different soils and her relatively simply and reliably from measurements of shear wave
results are given in Fig. 16. This shows the variations of the velocity in laboratory samples or in situ. The value of Go for a
parameters A, n and m with plasticity for ne grained soils. particular soil varies with current state in a simple and consis-
Although there is some scatter of the data there are clear trends tent manner given by equation (16) in which A, n and m are
showing that A decreases and both n and m increase with material parameters. For soils which are not strongly bonded or
increasing plasticity index. highly structured, these parameters depend principally on the
Coop and Jovicic (1999) reported the results of bender nature of the grains and vary consistently with plasticity index.
element tests on a variety of different coarse grained soils. They If the soil is assumed to be isotropic, the very small strain
found that the relationships between very small strain-stiffness Young's modulus Eo can be obtained from the shear modulus
Go and state given by equation (16) applied equally to coarse with an assumed value for Poisson's ratio. If the soil is cross-
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 495
3000 considered the measurement of Eo it is now necessary to
consider the measurement of the appropriate strength. It is well
known that a particular soil will have a number of different
strengths, depending on the drainage and the strain. It is
important to consider which strength is appropriate to determine
2000 rigidity and degree of non-linearity for routine designs which
Coefficient, A

are intended to limit movements or settlements.


Figure 17 illustrates the general features of the stressstrain
behaviour of a relatively stiff soil. There is a peak strength, an
ultimate or critical state strength and a residual strength. These
1000 may be described by undrained strengths su or by angles of
shearing resistance 9.
For complex numerical analyses, constitutive soil models
should include the complete stressstrain behaviour and might
include coupled loading and drainage. For simple routine de-
0 sign, it is necessary to choose either a drained or an undrained
strength and to choose between one or other of the peak,
(a )
ultimate or residual strengths. The choice between drained and
undrained strength depends on the relative rates of drainage and
loading but the choice between peak, ultimate and residual
strength is not so simple.
09
Figure 18 illustrates three typical cases. In Fig. 18(a) driving
a pile or landsliding has caused large relative displacements and
Coefficient, n

Shear
stress

07 Peak
Ultimate = critical state

Residual

05
(b)
Distortion
04

Fig. 17. Strength of soil

03
Coefficient, m

02

01
0 10 20 30 40 50 60 70
Plasticity index
(c)

Fig. 16. Material parameters for Go (Viggiani & Atkinson, 1995)

anisotropic and undrained, the Young's moduli and Poisson's


ratios can be found from Guhh and the degree of undrained
anisotropy N. Very small strain-stiffness, Eo , is one of the basic
parameters required to characterize rigidity and the degree of
non-linearity.

CHOICE OF SOIL STRENGTH FOR DESIGN


The parameters that, together with the very small strain-
stiffness Eo , characterize non-linearity are the strength qf and
the strain at failure f . These are used to describe the rigidity,
Eo =qf and the degree of non-linearity nl f =r, where r is Fig. 18. Choices of strength for design: (a) very large movements;
the reference strain which is the reciprocal of rigidity. Having (b) rst time failures; (c) small movements
496 ATKINSON
the creation of well-dened shear bands or slip surfaces. In ne inuence peak strengths and failure strains measured in soil
grained soil with platy particles this might reduce the strength samples.
to the residual which is then the strength which should be used
for subsequent shearing in the shear bands (Skempton, 1964).
For a rst time landslip in clay, illustrated in Fig. 18(b), the Angles of intersection of shear bands in undrained tests
movements are initially insufcient to reduce the strength to the Conventionally, directions and angles of intersection of shear
residual and the critical state strength should be used for design. bands in soil are associated with the stress eld. Shear bands
Back-analyses of shallow slips in motorway cuttings and em- are usually assumed to occur in the characteristic directions
bankment slopes (Crabb & Atkinson, 1991) showed that the which are the planes in the soil on which the stresses are those
strength mobilized at failure was very close to the ultimate or where the Mohr's circle touches the failure envelope: this leads
critical state strength: this is also very close to the fully to the standard Rankine stress elds in which stress character-
softened strength (Skempton, 1970). For coarse grained soils the istics and shear bands intersect at angles 90 .
critical state and residual strengths are the same. For undrained loading of soils this presents a problem
Figure 18(c) shows a loaded foundation with a settlement r. because, in terms of total stress, the failure envelope is given
If the bearing pressure is chosen on the basis of an ultimate by u 0 while, at the same time in terms of effective stress,
bearing capacity with a load factor, the strength used to it is given by a friction angle 9 which is non-zero. This is
calculate the bearing capacity should reect the stiffness of the illustrated in Fig. 20. The two Mohr's circles, one for total
soil. This cannot be the critical state strength. (If a foundation stress and one for effective stress, are for the same sample at
is designed on the basis of the critical state strength with the the same instant and the angle of intersection 2 of the stress
same load factor then the same foundation would be designed characteristics and the shear bands could be either 908 or
for dense and for loose sand which is clearly unsatisfactory.) 908 9. The question is: what are the directions of shear
Soil stiffness is related principally to peak strength and not to bands in undrained tests and, if shear bands are associated with
ultimate or critical state strength; stiffness and peak strength stress characteristics, how does the soil know which Mohr's
both increase with effective stress and overconsolidation. Conse- circle and failure criterion, in total or effective stress, to follow?
quently, it is the peak strength and the corresponding strain
which, together with Eo , should be used to determine rigidity
and degree of non-linearity for analyses which are intended to Formation of shear bands in triaxial tests
limit movements. Figure 21 shows two initially identical samples of overconso-
lidated kaolin clay after about 15% axial strain in unconned
compression tests. In both cases the overall water contents
SHEAR BANDS IN SOIL SAMPLES remained unchanged. The sample on the left developed a strong
Figure 19 shows a sample of reconstituted overconsolidated shear band. The shear band is inclined at about 608 to the
kaolin clay which has developed clearly dened discontinuities horizontal so complementary shear bands would intersect at
during a triaxial compression test. These discontinuities are about 608. The sample on the right barrelled; there is no distinct
usually called slip planes but, because they have nite thickness shear band.
(although perhaps only a few grains thick), they are known The difference between the two samples is that the one on
more properly as shear bands. It is quite obvious that once the left which had developed the shear band had been strained
distinct shear bands have developed they have a profound inu- very slowly and had reached 15% axial strain in about six hours
ence on the overall behaviour of soil in a sample in a laboratory while the one on the right which had barrelled had been
test and in the ground. strained very quickly and had reached 15% axial strain in about
The occurrence of slip planes or shear bands in soils is a one second. The shear bands which developed in the relatively
central feature of geotechnical stability analyses. There is slow undrained tests were in reconstituted samples and they
extensive literature on the initiation and development of shear appear also in coarse grained soil. They are not associated with
bands in soils and other granular materials, including theoretical pre-existing ssures.
studies (Vardoulakis & Sulem, 1995) and experimental observa- In ne grained soils, rate processes are associated primarily
tions (Viggiani et al., 1994; Finno et al., 1997). In his Rankine with drainage. In the sample that was strained very quickly it
Lecture, Burland (1990) analysed the behaviour of soil samples can be assumed that there was no drainage whatsoever and the
containing shear bands. It is, however, very difcult to obtain sample remained at constant volume throughout. In the rela-
meaningful soil parameters from such distorted and discontin- tively slow test, however, there is the possibility that there was
uous samples. Certainly there is very little that can be learned some local drainage near the developing shear band. Even if the
about soil stiffness once a shear band has developed in a permeability is low, the drainage path length may be only a few
sample. For the present, the question is how shear bands grain diameters long.

Volumetric straining and shear bands


In order to examine the directions of shear bands and volume
changes due to local drainage, it is necessary to consider
strains, not stresses. If the strains in material either side of a
shear band are small compared with the strains inside the shear
band then the material in a shear band will deform in plane
strain and the directions of shear bands will coincide with the
directions of zero extension (Roscoe, 1970).
Figure 22 shows Mohr's circles of strain increment for
increments of strain 1 and 3 . In the circle in Fig. 22(b)
there is no volumetric strain (because the centre of the Mohr's
circle is at the origin of the strain axes) and shear bands, which
coincide with directions of zero extension, intersect at 908. In
the Mohr's circle in Fig. 22(c) there is a volumetric strain v
and shear bands intersect at an angle 908 , where is the
angle of dilation.
This provides an explanation for the appearance and non-
appearance of shear bands in initially identical samples sheared
Fig. 19. Shear bands in a soil sample nominally undrained but at very different rates of strain. In Fig.
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 497

Fig. 20. Stress characteristics for undrained loading: (a) directions of stress characteristics; (b) Mohr's circle for total
stress; (c) Mohr's circle for effective stress

very different times to failure. They concluded that the angle


between shear bands is governed by the angle of dilation and
not by the angle of friction 9 or u . They showed that as the
time to failure increased, allowing more opportunity for local
drainage to occur and the angle of dilation to increase, the
observed angles of intersection of shear bands decreased. They
also found that if the test was carried out very quickly the
sample barrelled and shear bands did not develop.

Dilation in shear bands


Volume changes in shear bands have been observed directly
in laboratory tests, in model tests and in situ. Desrues et al.
(1996) observed voids ratios in shear bands in sand samples in
triaxial tests. They found that material within a developing shear
Fig. 21. Samples of kaolin clay after unconned compression tests band dilated and ultimately reached a unique critical state.
Figure 23 is an X-radiograph from the Rankine Lecture given
by Roscoe (1970). It shows passive loading of a wall in dry
sand as the wall moves from left to right. The tone of the image
is a negative of the voids ratio in the sand; the dark bands are
21, the sample on the left was strained relatively slowly; there
looser material. There are distinct bands of looser dilated sand.
was time for local drainage, soil in the shear band dilated,
These intersect at an angle which is about 608. Fig. 24 shows
softened and weakened and this allowed the shear band to grow.
water contents measured by Henkel (1956) across a shear band
The direction of the shear band in the sample in Fig. 21
in soil behind a failing retaining wall. The shear band is only a
corresponds to an angle of dilation of about 308. The sample
few tenths of an inch thick but the water content has increased
on the right was strained very quickly; there was no time even
by about 10%. These two sets of data show clearly that shear
for local drainage, there was no softening or weakening and no
bands have nite thickness in which soil has dilated and
strong shear band developed.
weakened.
Vardoulakis (1996a, 1996b) showed theoretically that shear
bands are associated with local volumetric straining and that
this can occur in globally undrained tests. He also showed that
if there is no volumetric straining in a shear band the resulting Effects of shear bands on peak strength in undrained tests
pore pressure gradients cannot be sustained in ordinary, slow, The inuence of rate of strain and local drainage on the
undrained tests. stressstrain behaviour and peak strength observed in nominally
Atkinson & Richardson (1987) measured the angles of inter- undrained triaxial tests was examined by Atkinson & Richard-
section of shear bands in initially identical samples of reconsti- son (1987). They showed that the directions of shear bands, the
tuted London Clay in nominally undrained triaxial tests with observed stressstrain behaviour, stress paths and peak strengths
498 ATKINSON

1

Pole
3 3

3 1

1 (b)

(a)



3 1

(c)

Fig. 22. Strains in shear bands: (a) directions of zero extension; (b) Mohr's circle for constant volume straining;
(c) Mohr's circle for straining with dilation

020 025 030 035


Water
content
4

Fig. 23. Dilation in shear bands in dry sand behind a model wall
(after Roscoe, 1970)
Distance
from soft Slip
zone in zone
observed in nominally undrained triaxial tests were all consis- inches
tent with local drainage and dilation in the shear bands.
Figure 25 shows the interpretation proposed by Atkinson &
Richardson (1987). Figs 25(a) and (c) show state paths with
axes q9, p9 and water content for overconsolidated samples with
and without local drainage. Fig. 25(b) shows the corresponding 2
stressstrain behaviour. The path ABF is fully undrained; there
is no volume change, the sample fails at F and there is no peak
strength. For the path CDE local drainage starts at D; ultimate
failure is at E, there is a peak strength at D which is the start
of local dilation. (Notice that the path ABF does have a peak
stress ratio and a peak 9 but not a peak deviator stress.)
The peak state at B corresponds to initiation of a shear band 4
owing to local drainage in a slow undrained test. In a fast
undrained test, in which there was no local drainage and no
shear band, the peak strength corresponds to the ultimate Fig. 24. Water content observed in a shear band in the ground
strength at F. For intermediate rates of loading, or times to behind a failing wall (after Henkel, 1956)
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 499
expectation that a shear band would miss one or more of the
local gauges. Fig. 26(b) shows strains measured by each of the
four local gauges plotted against the mean axial strain from a
test on overconsolidated kaolin clay. At a mean strain of about
5% one of the local gauges started to register strains signi-
cantly smaller than the others. This is taken to indicate the
initiation of a shear band.

25

20

Fig. 25. Behaviour of a sample with a dilating shear band (after


Atkinson & Richardson, 1987)
Local strain: %

15

failure, the peak strength will be between D and F. Atkinson & 10


Richardson (1987) found that the peak strength in nominally
undrained triaxial tests decreased with the logarithm of the time
to failure, which is consistent with the interpretation based on
local drainage in shear bands. 5

INITIATION OF SHEAR BANDS IN SOILS


0
The issue of localization and initiation of shear bands in 0 5 10 15 20
granular materials has been extensively researched theoretically Overall strain: %
(e.g. Vardoulakis & Sulem, 1995) and experimentally (e.g.
Finno et al., 1997). At City University, Albert (1999) carried Fig. 26. Initiation of a shear band in a triaxial sample: (a)
out triaxial tests on samples of a variety of different soils using equipment; (b) Local strains in a sample of overconsolidated kaolin
four local axial-strain gauges, as shown in Fig. 26(a), in the clay
500 ATKINSON
Figure 27 shows the stress path and stressstrain curves for assumed to be undrained overall. Consequently, tests to measure
the test on overconsolidated kaolin shown in Fig. 26. The point peak strength and failure strain, which will be used to determine
of initiation of the shear band at an axial strain of 5%, as rigidity and the degree of non-linearity, should be carried out
interpreted from Fig. 26(b), is indicated by the arrows. These sufciently slowly to allow shear bands to develop. Peak
data show that the start of non-uniform straining, and the strengths measured in samples that develop shear bands will be
initiation of a shear band, occur a little before the peak deviator smaller than those measured in samples that do not develop
stress and a little after the peak stress ratio. Similar results were shear bands.
obtained by Albert (1999) on other soils and have also been
obtained by several others (Viggiani et al., 1994).
TYPICAL VALUES FOR SOIL NON-LINEAR PARAMETERS
Soils are essentially collections of grains with grading,
PEAK STRENGTH, FAILURE STRAIN AND SHEAR BANDS mineralogy, shape and texture that are arranged in a dense or
Peak strength, qf , and failure strain, f , together with stiff- loose packing and are loaded to a current effective stress. In the
ness at very small strain Eo are the parameters which are absence of strong bonding or structure, the parameters which
needed for the determination of rigidity and the degree of non- describe the mechanical behaviour of a soil should depend
linearity. There is evidence that the peak strength of a soil is primarily on its nature and its state and they should generally
associated strongly with local drainage and dilation and with fall within certain limits. Some parameters, such as critical state
the initiation of a shear band. It is the local deformations and friction angle 9c , are material parameters and they depend only
volume changes leading to the development of a shear band on the nature of the grains. Others, such as very small strain
which primarily determine peak strength and failure strain. shear modulus Go , depend also on the state. Other parameters,
In most practical cases, rates of construction are sufciently such as A, n and m in equation (16), that relate mechanical
slow to allow shear bands to develop in the ground if the properties to state, are themselves material parameters and are
structure comes close to failure even if the ground can be related to the plasticity index as shown in Fig. 16.

Typical values for stiffness and strength of soils


150
Tatsuoka & Shibuya (1992) collected data of compressive
strength qf and very small strain Young's modulus Eo for many
different soils in many different states. They also included data
for other materials, including rocks, concrete and metals. Their
100
data for typical soils fall within the shaded region in Fig. 28.
The range of compressive strength qf is about 80 kPa to
400 kPa; the range of Eo is about 80 MPa to 600 MPa; the
q : kPa

range of rigidity Eo =qf is about 3000 to 500. These data do


50
not, however, distinguish between different soils and between
different states of the same soil.

0
Variation of rigidity with plasticity and state
(a) The stiffness of soil at very small strain is related to its state
15 by equation (16) in which the parameters A, n and m depend
on the plasticity index as shown in Fig. 16. For undrained
loading of isotropic elastic soil Euo 3Go and equation (16)
becomes
10  n
Euo p9
3A Rom (17)
q /p

pa pa

05

0 Eo/q f
0 5 10 15 20 104 103
a: % 106
(b)
150

100
Eo: kPa

102
q : kPa

105

50

0
0 50 100 150 104
p : kPa 10 102 103
(c) q f: kPa

Fig. 27. Initiation of a shear band in a sample of overconsolidated Fig. 28. Typical values of strength and stiffness of soils (after
kaolin clay Tatsuoka & Shibuya, 1992)
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 501
The undrained compressive strength qf is also related to the Variation of failure strain with state
state by a relationship of the form (Muir Wood, 1990) Figures 30(a) and (b) show variations of strain at failure, at
  the peak deviator stress, with initial state for a number of
qf p9 different soils for drained and undrained triaxial tests. With the
2B R (18)
pa pa o exception of the data for Brasted Sand (Cornforth, 1967) the
data are from tests carried out at City University. These data
For normally consolidated soil Ro 1 and equation (18) is show that, in general, failure strain reduces with increasing
equivalent to su = v9 B, where su is the undrained strength, v9 overconsolidation ratio and with decreasing specic volume.
is the vertical effective stress and B is related to the plasticity This means that, in general, failure strain, and hence degree of
index (Skempton, 1957). For overconsolidated soils, the param- non-linearity, tends to decrease as the state moves away from a
ter is approximately 08 for a wide range of soils (Muir reference line as indicated in Fig. 30(c) (that is the degree of
Wood, 1990). non-linearity will tend to decrease with overconsolidation).
Dividing equation (17) by equation (18) gives The data shown in Figs 29 and 30 demonstrate that, at least
  for reconstituted soils, rigidity and the degree of non-linearity
Euo A p9 n1 m vary consistently with the nature of the soil grains and with the
1:5 Ro (19)
qf B pa current state of the soil. Engineers like to believe that engineer-
ing properties of soils are variable. Indeed they are but they do
in which the rigidity (Eo =qf ) is related to the current state vary in a consistent and predictable way.
(given by p9 and Ro ) through material parameters which
themselves depend on the plasticity index.
Figure 29 shows values of rigidity given by equation (19), INFLUENCE OF SOIL NON-LINEARITY ON DESIGN
varying with current pressure and with the overconsolidation PARAMETERS
ratio for soils with different plasticity indices. To evaluate Non-linearity in soil can be described by rigidity and the
equation (19), values of A, n and m were taken from Fig. 16; degree of non-linearity and it is interesting to examine how
values for B and were taken from Muir Wood (1990). The these inuence choices of parameters for simple routine design.
range of values of rigidity in Fig. 29 is about 400 to 4000, To do this it is helpful to make use of a simple expression for
which is only a little larger than that given in the data by non-linear stressstrain behaviour.
Tatsuoka & Shibuya (1992) and shown in Fig. 28. There are many expressions for non-linear stressstrain
Figure 29(a) shows that the rigidity of soil decreases with curves for soil in the literature (e.g. Kondner, 1963; Puzrin &
stress (for a given overconsolidation ratio). This is because the Burland, 1998). The expression in equation (20) is about the
value of n is always smaller than 1 and so, in equation (19), simplest that captures the essential features of non-linear
n 1 is always negative. Fig. 29(b) shows that the rigidity of stressstrain behaviour.
soil decreases with the overconsolidation ratio (for a given  r
stress). This is because the value of m is always smaller than f
1
the value of and so m is always negative. These results Et
 r (20)
explain why the rigidity of stiff soil is smaller than the rigidity Eo f
of soft soil, as noted in Table 1. 1
o
The tangent Young's modulus Et decays with strain; there is a
Typical values for o region of very small strain where Et Eo up to a limiting
The limiting strain within which the stiffness of soil may be strain o ; there is a failure strain f . There is also a compressive
taken to be constant with a value Go or Eo is o . It may be strength qf which xes the value of r so that the area beneath
observed in resonant column tests (Georgiannou et al., 1991) or the stiffnessstrain curve is qf . (For typical values of rigidity
in triaxial tests using precise local gauges (Coop et al., 1997). and degree of non-linearity for soil the value of r is generally
From results of resonant column tests, the limiting shear in the range 01 to 05.)
strain o ( 2o ) was found to increase with the plasticity index Equation (20) is applicable to drained or to undrained load-
from about 103 % (o 0:0005%) for low plasticity silts to ing, with appropriate values for the parameters. It can be
about 102 % (o 0:005%) for high plasticity clays integrated to give a simple expression q q() and the secant
(Georgiannou et al., 1991). From results of triaxial tests, with Young's modulus can be calculated from this. By varying the
very precise local gauges, Coop et al. (1997) found o smaller parameters Eo , qf and f , this describes the stressstrain
than 00001% for unbonded coarse grained soils. For bonded behaviour of soils with different rigidities and different degrees
soils and soft rocks, values of o are relatively large; Cuccovillo of non-linearity.
& Coop (1997) found o about 002% in tests on intact samples Figure 31 shows soil behaviour given by equation (20)
of Greensand. plotted for soil with a rigidity of 1000 and for degrees of non-

Fig. 29. Variation of rigidity with state and plasticity index


502 ATKINSON

Fig. 30. Variation of failure strain with state: (a) ne grained soils; (b) coarse grained soils; (c) general features

linearity in the range 10 to 100. In Fig. 31(a) the load factor 12 n l = 100
Lf q=qf . The data illustrate how load factor, Lf , and secant
stiffness ratio, Es =Eo , vary with the degree of non-linearity. At n l = 50
a strain of 01%, indicated by the arrows, both load factor and
stiffness ratio vary by factors of 2 to 3. n l = 20
Figures 32(a) and (b) show the variations of load factor and 08
stiffness ratio with the degree of non-linearity and rigidity for a n l = 10
strain of 01%. Over a typical range of non-linear soil parameters
Lf

the stiffness ratio needed for a design strain of 01% varies from Eo /q f = 1000
about 05 to less than 02 and the variation of load factor is
greater. Fig. 32(c) shows how the ratio of stiffness to strength 04
(Es =su ) varies with non-linearity; this is a parameter often used
in simple routine design. Similar design curves can be easily
developed for other characteristic strains from a simple stress
strain equation, such as that given in equation (20).
The data given in Figs 31 and 32 were calculated for a value 0
of o 0:001%. For smaller values o has little inuence on the (a)
load factor or stiffness ratio. For bonded soils and soft rocks, 12
however, o may be considerably larger than 0001% and then
the value of o begins to have an inuence on the load factor
and stiffness ratio. For these materials, a better basis for design
may be to avoid strains in the ground that are greater than the
value of o , especially if the material is brittle with a rapid drop 08
of stiffness with strain after o .
Es/Eo

STRAINS IN TRIAXIAL SAMPLES AND GROUND MOVEMENTS


The data in Figs 31 and 32 relate load factor and stiffness 04
ratio to non-linear parameters for strains in a triaxial test speci-
men. But routine designs consider ground movements and so it
is necessary to relate strains in triaxial specimens to ground
movements.
Figure 33(a) shows the results of nite element calculations 0
for a triaxial sample and for shallow strip and circular founda- 103 102 101 1 10 102
tions all for undrained loading (Simpson, 2000). All the : %
(b)
analyses were carried out using the BRICK model (Simpson,
1992) with the same set of material parameters. The stress Fig. 31. Variation of design parameters with strain and degree of
strain curve for the triaxial sample is plotted with axes q and non-linearity
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 503
10 From the loadsettlement curves in Fig. 33(a), values of
Eo /q f = 1500 secant Young's modulus were calculated from
= 01%

r (1 2 )
Eo /q f = 1000 Ir (21)
B Es
Eo /q f = 500
where Ir is an appropriate inuence factor (Poulos & Davis,
Lf

05
1974) and, for undrained loading, u 12. Values for the un-
drained secant Young's modulus Eus , calculated from equation
(21) from the loadsettlement curves for shallow circular and
strip foundations in Fig. 33(a), are shown in Fig. 33(b) plotted
against the settlement to width ratio r=B. Also shown in Fig.
0 33(b) are values for the undrained secant Young's modulus Eus
(a) for a triaxial sample calculated from the triaxial stressstrain
10
curve in Fig. 33(a).
From Fig. 33(b) the values of r=B for a shallow foundation
are two to three times larger than the axial strains in a triaxial
sample at the same average stiffness. These results mean that
the stiffness at a certain strain measured in a triaxial specimen
relates to the design stiffness for a foundation at values of r=B,
E s /E o

05
which are two to three times larger than the corresponding axial
strain in the triaxial sample. Bolton (1993) has obtained similar
results for shallow and deep foundations and for retaining walls
using plasticity analyses.

0
NON-LINEARITY IN MODEL AND FULL-SCALE FOUNDATIONS
(b)
The inuence of non-linearity on foundation behaviour and
2000
on the choice of design stiffness can be illustrated by relating
the non-linear load settlement behaviour of model and full-scale
foundations to the non-linear characteristics of the soil.

Foundations in London Clay


Es/su

1000
Figure 34 shows stiffnessstrain data obtained from observa-
tions of the settlements of shallow and piled raft foundations on
London Clay made by Arup Geotechnics (1991) together with
the corresponding behaviour of London Clay in a triaxial test.
For each foundation case record the equivalent undrained secant
0 Young's modulus Eus was calculated from the bearing pressure
1 10
nl
102 103 and from the observed settlement using equation (21). In Fig.
(c) 34(a) the data are plotted as Eus =Euo and in Fig. 34(b) they are
plotted as Eus =su , which is often used to choose a value for
Fig. 32. Variation of design parameters with rigidity and degree of stiffness for routine design. Values for Euo and su were estimated
non-linearity for 0:1% from the site investigation data for each site.
Also shown in Figs 34(a) and (b) are broken lines that
represent the behaviour of London Clay in an undrained triaxial
a . The behaviour of the foundations is plotted as bearing test. The data for these were calculated using the simple model
pressure and settlement to width ratio r=B. All three curves given in equation (20) with parameters for London Clay which
approach constant stress at relatively large strains or settlements were a best estimate for the mean values for the many sites
and the bearing capacity is linked to the compressive strength considered. (The procedure used was to integrate equation (20),
through an appropriate bearing capacity factor. select a value of r to give the required values of qf and obtain

100
40
Strip foundation

80 Circular foundation
30 Triaxial test
q or : kPa

60
Esu: MPa

20
40

10
20

0
0 05 10 15 20 25 103 102 101 1 10
/B or a: % /B or a: %
(a) (b)

Fig. 33. Finite element analyses of shallow foundations and a triaxial test for the same soil
504 ATKINSON
10
Shallow rafts

Piled rafts
08
Triaxial test

06 Triaxial strains 3
Esu/Eou

04

02

0
(a)
2000

1500 Soil parameters:


Eou/q f = 1000
n l = 50
Esu/su

1000

500

0
103 102 101 1 10
/B or a: %
(b)

Fig. 34. Settlement of foundations on London Clay (data from Arup


Geotechnics, 1991)

values of the secant Young's modulus from the calculated Fig. 35(a). London Geotechnical Centrifuge. (b) Centrifuge model
stressstrain curve.) The solid lines in Fig. 34 are the lines for foundation
the triaxial test with strains increased three times to account for
the differences between axial strains in triaxial samples and
values of r=B for foundations.
For the foundations, the stiffnesses back-calculated from the (Stallebrass & Taylor, 1997). The foundation was 60 mm dia.
eld observations decay with increasing values of r=B in the and during the test the centrifuge acceleration was 100 g so the
same way that stiffness decays with strain in a triaxial test. experiment was modelling a foundation 6 m dia. The instru-
The values are, however, smaller than those corresponding to ments measured the prole of surface settlement but only the
the line for the triaxial test with strains increased by three settlement of the rigid foundation will be considered here.
times. This is thought to be due to some drainage that probably Data from a foundation loading test are shown in Fig. 36.
occurred in the ground during construction and foundation
loading. Drainage would have the effect of increasing settle-
ments and so reducing calculated stiffnesses.
The values of r=B observed for foundations on London Clay
10
shown in Fig. 34 are in the range of about 005% to 05%.
These are comparable to the typical strain range for foundations Triaxial test
given by Mair (1993) and shown in Fig. 1. Triaxial strains 3
08
Model test
Centrifuge model foundation on kaolin clay
Although it is always valuable to be able to compare theor- 06
etical analyses with full-scale observations, it is often difcult
Esu/Eou

to obtain all the required information about the soil, the Soil parameters:
structure, its loads and settlements and the drainage conditions. 04 Eou/q f = 5000
Many of these uncertainties are avoided by observation of the n l = 200
behaviour of closely monitored scale models using well docu-
mented soils. Since soils are essentially frictional materials and 02
many of their stiffness and strength properties depend on the
current effective stress, geotechnical models should correctly
scale effective stress. Effective stress scaling can be achieved by 0
testing models in a geotechnical centrifuge (Schoeld, 1980). 103 102 101 1 10
Figure 35(a) shows the London Geotechnical Centrifuge at /B or a: %
City University (Schoeld & Taylor, 1988) and Fig. 35(b) shows
a detail of a scale-model rigid shallow foundation on over- Fig. 36. Settlement of a centrifuge model foundation on kaolin clay
consolidated kaolin clay which was loaded in the centrifuge (data from Stallebrass & Taylor, 1997)
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 505
The data are shown in the same form as the data for the For both sands the secant stiffnesses back-calculated from the
foundations on London Clay in Fig. 34(a). As before, the lines model plate tests from equation (21) decay with settlement and
for a triaxial test were calculated from equation (20) with in both cases they are close to those given by the lines for
parameters that were best estimates for the kaolin clay at the triaxial tests with strains increased by three times.
states near the model foundation.
The stiffnesses back-calculated from the load-settlement be-
haviour of the model foundation decay with increasing settle- INFLUENCE OF NON-LINEARITY ON STIFFNESS RATIO FOR
ment. These stiffnesses are close to those given by the line for DESIGN
a triaxial test with strains increased by three times. Figure 38 shows values for the drained and undrained secant
Young's modulus, back-calculated from model tests of founda-
tions and plates on silica sand, carbonate sand and kaolin clay.
Model plate loading tests in sands For the same value of r=B, the stiffness ratio is signicantly
Alternatively, model tests may be carried out at elevated larger for silica sand than for carbonate sand or kaolin clay, or,
effective stress in a calibration chamber. In this case stresses are for the same stiffness ratio, the settlements for a foundation on
assumed to be uniform with depth rather than increasing with silica sand would be signicantly larger than the settlements of
depth as in a centrifuge model. foundations on carbonate sand or kaolin clay. These differences
Figure 37 shows data from loading tests on model plates in are due principally to the different degrees of non-linearity and,
dry carbonate sand and in dry silica sand in a calibration to a lesser extent, to the different rigidities of the three
chamber (Jamiolkowski, 2000). The initial mean-effective stress materials.
in these tests was generally in the range of 50 kPa to 180 kPa. The model plates did not fail; in each case the bearing
The data are shown in the same form as the data for the pressure continued to increase even after very large settlements.
foundations on London Clay in Fig. 34(a), except that the Consequently it is difcult to identify a bearing capacity and to
secant Young's moduli E9s are now in terms of effective stress investigate the relationship between load factor, settlement and
corresponding to drained loading of dry sand. soil characteristics.
The lines for the triaxial tests were calculated from equation
(20) with parameters that were best estimates for the effective
stress parameters for the two sands at the initial states near the SIMPLE DESIGN PROCESS
model plates in the calibration chamber. The rigidity taken for Design is always an iterative process and Fig. 39 illustrates a
the carbonate sand ( 1500) was the same as that taken for the simple method for routine design to take account of non-
silica sand while the degree of non-linearity taken for the linearity.
carbonate sand (nl 150) is very much larger than the degree The essence is to obtain a relationship between stiffness ratio
of non-linearity taken for the silica sand (nl 15). This reects Es =Eo and r=B. This requires: measurement of very small
the observation that the strain at failure at the peak strength in strain stiffness Eo , strength qf and failure strain f ; construction
carbonate sand is often relatively large while for silica sand at
the same state it is usually relatively small.

Fig. 38. Settlements of model foundations on sands and clay

Measure Eoq t and f Draw stiffnessstrain curves


Es /Eo

/B 3

Triaxial

log or /B

Select and B Find bearing pressure

Determine Es /B = (1 2)/(Es)l
Fig. 37. Settlements of model plates: (a) carbonate sand; (b) silica
sand (data from Jamiolkowski, 2000) Fig. 39. Simple design process
506 ATKINSON
of a triaxial stressstrain curve from equation (20) or a similar strongly bonded or highly structured, or if full stress, strain and
relationship; construction of a curve of stiffness ratio Es =Eo displacement elds are needed, more complex procedures will
against r=B using a relationship between triaxial strain and be required.
ground movements. For a foundation, values of r=B are about
three times the axial strain in a triaxial test for the same
stiffness. ACKNOWLEDGEMENTS
The design process then iterates around a loop as illustrated, I am indebted to many people who contributed in one way or
until the loads, dimensions, stiffnesses and settlements are all another to the lecture and to the written paper. I am very
compatible. fortunate to have worked for nearly 20 years in a dynamic and
intellectually demanding research group at City University. I
owe much to Professor Raoul Franklin, himself originally a civil
SUMMARY AND CONCLUSIONS engineer, who was Vice-Chancellor for much of that time and
The stressstrain behaviour of soil is highly non-linear and who had the foresight to establish and support research centres
stiffness for both drained and undrained loading decays with in the University. Much of the work which was the foundation
strain. This has implications both for testing, to determine for the lecture was done by research students, research assis-
design parameters, and for calculations of ground and structure tants, technicians and visitors at City University. During the
movements. Full analyses of geotechnical structures require preparation of the lecture and paper, my colleagues, Neil Taylor,
special laboratory tests and complex numerical calculations. Matthew Coop and Sarah Stallebrass, shielded me from many
Alternatively, simpler routine analyses which calculate move- distractions and provided me with information, data and encour-
ment in only one direction require load factors or stiffnesses agement. They and others at City University heard several
which, owing to soil non-linearity, depend on movements and rehearsals of the lecture and commented on the paper. Collea-
strains. gues at Arup Geotechnics also heard rehearsals of the lecture
Simple non-linear stressstrain behaviour can be character- and helped with calculations. I am particularly grateful to
ized by rigidity, Eo =qf , and by degree of non-linearity, f =r , Gioacchino Viggiani for help with shear bands, to Martin Lings
where the reference strain, r , is the reciprocal of rigidity. These for help with anisotropy, to Tony Butcher and Mike Jamiolk-
soil parameters can be measured in routine and relatively simple owski for test data, to Vojkan Jovicic for photographs and other
tests. help and to Brian Simpson and Sarah Stallebrass for nite
Young's modulus at very small strain, Eo , can be obtained element calculations. Finally, I am grateful to my family, Jo,
from measurements of shear wave velocity in situ from down- Robert and Nicholas; their job was to get me to the lecture on
hole or cross-hole tests or in laboratory samples using bender time and well prepared, and this they did. Nicholas made a
elements. For isotropic soil, relationships between shear wave video which was an important part of the lecture and he also
velocity and stiffness parameters are simple but for anisotropic made many of the diagrams.
soil they are more complicated and for full interpretations
additional tests are required. Eo is an important design para-
meter as it contributes towards characterizing non-linear beha- REFERENCES
viour and it is the basis for stiffness ratio Es =Eo for routine Abbiss, C. P. & Ashby, K. D. (1983). Determination of ground moduli
design. by a seismic noise technique on land and on the sea bed. Geotechni-
The strength and failure strains which characterize non- que 33, No. 4, 445451.
linearity in soil are those at the peak state. They can be Albert, C. (1999). Personal communication.
measured in conventional laboratory tests but they are strongly Arup Geotechnics. (1991). End of construction (undrained) settlement of
dependent on local drainage and the initiation of shear bands. structures on London Clay. Report No. 5294=57.
Atkinson, J. H. (1973). The deformation of undisturbed London Clay.
Many soil stiffness and strength parameters depend on cur- PhD thesis, University of London.
rent state, measured as the distance of the state point from a Atkinson, J. H. (1975). Anisotropic elastic deformations in laboratory
reference line. For reconstituted soils and for natural soils, that tests on undisturbed London Clay. Geotechnique 25, No. 2, 357
are not strongly bonded or structured, stiffness and strength 374.
parameters depend on the nature and on the state of the soil in Atkinson, J. H. & Evans, J. S. (1985). Discussion on: The measurement
relatively simple ways. Stiffness at very small strain is related of soil stiffness in the triaxial apparatus by R. J. Jardine, M. J.
to current state through parameters that themselves depend Symes, and J. B. Burland, Geotechnique 35, No. 3, 378382.
principally on the nature of the soil. The strain at failure at the Atkinson, J. H. & Richardson, D. (1987). The effect of local drainage in
peak state also varies with state. shear zones on the undrained strength of overconsolidated clay.
Geotechnique 37, No. 3, 393403.
Soil stiffness parameters obtained from back-analyses of Atkinson, J. H. & Sallfors, G. (1991). Experimental determination of
observed settlements of full-scale and model foundations are soil properties. General Report to Session 1. Proceedings of the 10th
non-linear and decay with settlement in the same way that ECSMFE, Florence 3, 915956.
stiffness decays with strain in a triaxial test sample. For equiva- Baldi, G., Hight, D. W. & Thomas, G. E. (1988). A re-evaluation of
lent stiffnesses, values of the ratio of foundation settlement to conventional triaxial test methods. In Advanced triaxial testing of
width (r=B) are two to three times larger than the correspond- soil and rock. ASTM. STP 977, pp 219263.
ing axial strains in a triaxial test. By taking this difference into Been, K. & Jefferies, M. G. (1986). A state parameter for sands.
account, stiffnesssettlement relationships back-calculated from Geotechnique 36, No. 1, 127132.
Bishop, A. W. & Henkel, D. J. (1957). The measurement of soil proper-
observed settlements of model foundations and plates are con-
ties in the triaxial test. Edward Arnold.
sistent with the non-linear stressstrain behaviour measured in Bolton, M. D. (1993). Design methods. Proc. Wroth Mem. Symp:
simple laboratory tests. Stiffness back-calculated from observed Predictive Soil Mechanics. 5071.
settlements of foundations in London Clay decayed with settle- Burland, J. B. (1979). Contribution to discussion on Session 4. Proc 7th
ment in the same way but were smaller than those correspond- European Conf. SMFE, Brighton, 4, 137.
ing to stiffnessstrain relationships measured in triaxial tests. Burland, J. B. (1990). 30th Rankine Lecture. On the shear strength and
The differences are attributable to partial drainage in the ground compressibility of natural clays. Geotechnique 40, No. 3, 329378.
during construction. Butcher, A. P. (2000). Private communication.
Non-linear soil behaviour can be taken into account in simple Campanella, R. G., Robertson, P. K. & Gillespie, D. (1986). Seismic
routine design procedures. The parameters which characterize cone penetration tests. Proc. In situ 86, Specialty Conf. on Use of In
Situ Tests in Geotech. Engng, Blacksburg, ASCE, 116130.
non-linear stiffness can be measured in simple routine tests and Chandler, R. J. (2000). 3rd Glossop Lecture: Clay sediments in deposi-
the analyses are straightforward. These simple methods are tional basins: the geotechnical cycle. QJEG 33, Part 1, 739.
however limited. They can determine movements in only one Cole, K. W. & Burland, J. B. (1972). Observation of retaining wall
direction, for relatively simple structures and for soils which are movements associated with a large excavation. Proc. 5th ECSMFE,
well behaved. For complex structures, or for soils which are Madrid.
NON-LINEAR SOIL STIFFNESS IN ROUTINE DESIGN 507
Coop, M. R. & Lee, I. K. (1993). The behaviour of granular soils at Schoeld, A. N. (1980). Cambridge geotechnical centrifuge operations.
elevated stresses. Proc. Wroth Mem. Symp: Predictive Soil Mech- Geotechnique 30, No. 3, 227268.
anics. 186198. Schoeld, A. N. & Wroth, C. P. (1968). Critical state soil mechanics.
Coop, M. R. & Jovicic, V. (1999). The inuence of state on the very McGraw-Hill.
small strain stiffness of soils. 2nd Int. Symp. Pre-failure Deformation Schoeld, A. N. & Taylor, R. N. (1988). Development of standard
of Geomaterials IS-Torino '99, Turin, 175181. geotechnical centrifuge operations. In Centrifuge 88 (ed. J.-F. Corte),
Coop, M. R., Jovicic, V. & Atkinson, J. H. (1997). Comparisons between Rotterdam: Balkema, pp. 2932.
soil stiffnesses in laboratory tests using dynamic and continuous Scholey, G. K., Frost, J. D., Lo Presti, D. C. F. & Jamiolkowski, M.
loading. Proc. 4th Int. Conf. SMFE, Hamburg 1, 267270. (1995). A review of instrumentation for measuring small strains
Cornforth, D. H. (1964). Some experiments on the inuence of strain during triaxial testing of soil specimens. Geotechnical Testing J 18,
conditions on the strength of Sand. Geotechnique 14, No. 2, 143167. No. 2, 137156.
Crabb, G. I. & Atkinson, J. H. (1991). Determination of soil strength Shirley, D. J. & Hampton, L. D. (1978). Shear wave measurements
parameters for the analysis of highway slope failures. In Slope in laboratory sediments. J. Acous. Soc. of America 63, No. 2,
stability engineering: developments and applications (ed. R. Chand- 607613.
ler). London: Thomas Telford, pp. 1318. Simpson, B. (1992). 32nd Rankine Lecture. Retaining structures: dis-
Cuccovillo, T. & Coop, M. R. (1997). The measurement of local axial strain placement and design. Geotechnique 42, No. 2, 149177.
in triaxial tests using LVDT's. Geotechnique 47, No. 1, 167171. Simpson, B. (2000). Personal communication.
Desrues, J., Chambon, R., Mokni, M. & Mazerolle, F. (1996). Void ratio Simpson, B., Atkinson, J. H. & Jovicic, V. (1996). The inuence of
evolution inside shear bands in triaxial sand specimens studied by anisotropy on calculations of ground settlements above tunnels.
computed tomography. Geotechnique 46, No. 3, 529546. Proc. Int. Symp on Geotechnical Aspects of Underground Construc-
Dyvik, R. & Madshus, C. (1985). Laboratory measurements of Gmax tion in Soft Ground. Rotterdam: Balkema, 591594.
using bender elements. ASCE convention: Advances in the art of Skempton, A. W. (1957). Discussion. The planning and design of the
testing soils under cyclic conditions, Detroit, Michigan. new Hong Kong airport. Proc. Instn Civ. Engrs 7, 305307.
Finno, R. J., Harris, W. W., Mooney, M. A. & Viggiani, G. (1997). Skempton, A. W. (1964). 4th Rankine Lecture. Long term stability of
Shear bands in plane strain compression of loose sand. Geotechnique clay slopes. Geotechnique 14, No. 2, 77101.
47, No. 1, 149165. Skempton, A. W. (1970). First time slides in over-consolidated clays.
Georgiannou, V. N., Rampello, S. & Silvestri, F. (1991). Static and Geotechnique 20, No. 3, 320324.
dynamic measurement of undrained stiffness of natural overconsoli- St John, H. D. (1975). Field and theoretical studies of behaviour of
dated clays. Proc. 10th ECSMFE, Florence 1, 9196. ground around deep excavations in London Clay. PhD thesis,
Gibson, R. E. (1974). 14th Rankine Lecture. The analytical method in University of Cambridge.
soil mechanics. Geotechnique 24, No. 2, 115140. Stallebrass, S. E. & Taylor, R. N. (1997). The development and
Hardin, B. O. (1978). The nature of stressstrain behaviour for soils. evaluation of a constitutive model for the prediction of ground
Proc. Geot. Div. Specialty Conf. on Earthquake Engng and Soil movements in overconsolidated clay. Geotechnique 47, No. 2, 235
Dynamics, Pasadena 1, 339. 253.
Hardin, B. O. & Drnevich, V. P. (1972). Shear modulus and damping in Tatsuoka, F. & Shibuya, S. (1992). Deformation characteristics of soils
soils. Measurement and parameter effects. J. of Geotech. Engng Div. and rocks from eld and laboratory tests. Proc 9th Asian Regional
ASCE 102, No. GT9, 975978. Conf. on SMFE 2, 101170.
Henkel, D. J. (1956). Discussion on: Earth movements affecting LTE Terzaghi, K. & Peck, R. B. (1948). Soil mechanics in engineering
railway in deep cutting east of Uxbridge. Proc. Instn Civ. Engrs, practice. John Wiley.
Part 2, 5, 320323. Vardoulakis, I. (1996a). Deformation of water saturated sand: I. uniform
Hight, D. W. (1998). 38th Rankine Lecture. Soil characterization: the undrained deformation and shear banding. Geotechnique 46, No. 3,
importance of structure, anisotropy and natural variability. Geotech- 441456.
nique (to be published). Vardoulakis, I. (1996b). Deformation of water saturated sand: II. effect
Hight, D. W. & Higgins, K. G. (1995). An approach to the prediction of of pore water ow and shear banding. Geotechnique 46, No. 3,
ground movements in engineering practice: background and applica- 457472.
tion. Proc. Int. Symp. on Pre-failure Deformations of Geomaterials, Vardoulakis, I. & Sulem, J. (1995). Bifurcation analysis in geomech-
IS-Hokkaido 2, 909945. anics. Blackie Academic and Professional.
Jamiolkowski, M. (2000). Personal communication. Vesic, A. S. (1972). Expansion of cavities in an expanding soil
Jardine, R. J., Symes, M. J. & Burland, J. B. (1984). The measurement mass. Proc. ASCE, J. Soil Mech. & Found. Engng Div. 98, SM3,
of soil stiffness in the triaxial apparatus. Geotechnique 34, No. 3, 265290.
323340. Viggiani, G. (1992). Small strain stiffness of ne grained soils. PhD
Jovicic, V. (1997). The measurement and interpretation of small strain thesis, City University, London.
stiffness of soils. PhD thesis, City University, London. Viggiani, G. & Atkinson, J. H. (1995). Stiffness of ne-grained soil at
Jovicic, V. & Coop, M. R. (1997). The stiffness of coarse grained soils very small strain. Geotechnique 45, No. 2, 249265.
at small strains. Geotechnique 47, No. 3, 545561. Viggiani, G., Finno, R. J. & Harris, W. W. (1994). Experimental
Jovicic, V., Coop, M. R. & Simic, M. (1996). Objective criteria for observations of strain localisation in plane strain compression of a
determining Gmax from bender element tests. Geotechnique 46, No. stiff clay. In Localisation and bifurcation theory for soils and rocks
2, 357362. (eds R. Chambon, J. Desrues & I. Vardoulakis). Rotterdam: Balk-
Kondner, R. L. (1963). Hyperbolic stressstrain response: cohesive ema, pp. 189198.
soils. J. Soil Mech. and Foundation Engng Div. ASCE 89, No. 1, Wroth, C. P. (1975). In situ measurement of initial stresses and
115143. deformation characteristics. Proc. Geot. Engng Div. Specialty Conf.
Lings, M. L., Pennington, D. S. & Nash, D. F. T. (2000). Anisotropic on In Situ Measurement of Soil Properties. ASCE 2, 181230.
stiffness parameters and their measurement in a stiff natural clay.
Geotechnique 50, No. 2, 109125.
Mair, R. J. (1993). Developments in geotechnical engineering research:
VOTE OF THANKS
applications to tunnels and deep excavations. Unwin Memorial
Lecture 1992. Proc. Instn Civ. Engrs Civ. Engng, 3, 2741. DR B. SIMPSON, Director, Arup Geotechnics, London and
Muir Wood, D. M. (1990). Soil behaviour and critical state soil mech- Honorary Editor of Geotechnique
anics. Cambridge University Press.
Pennington, D. S. (1999). The anisotropic small strain stiffness of The last time I had to speak from this rostrum, I was given
Cambridge Gault Clay. PhD thesis, University of Bristol. some friendly advice from Professor Peter Vaughan: `decide
Pennington, D. S., Nash, D. F. T. & Lings, M. L. (1997). Anisotropy of which of your slides you want the audience to understand and
Go shear stiffness in Gault Clay. Geotechnique 47, No. 3, 391398. which you do not want them to understand!' Sometimes lectures
Poulos, H. G. & Davis, E. H. (1974). Elastic solutions for soil and rock set out to bring clarity to some parts of their presentation, but
mechanics. John Wiley.
to obscure in others. Contrast John Atkinson: in the Lecture we
Puzrin, A. M. & Burland, J. B. (1998). Non-linear model of small-strain
behaviour of soils. Geotechnique 48, No. 2, 217233. have seen a typical example of John's workpart of his mission
Richart, F. E., Hall, J. R. & Woods, R. D. (1970). Vibration of soils and was to take relatively complex information and to clarify and
foundations. New Jersey: Prentice-Hall Inc. simplify it where possible, and to make it accessible for
Roscoe, K. H. (1970). 10th Rankine lecture. The inuence of strains in practical use.
soil mechanics. Geotechnique 20, No. 2, 129170. The more we know about soils the more we realize their
508 ATKINSON
complexity. John has referred to features such as non-linearity, However, John's aim in the Lecture has been to show how
anisotropy, shear bands, dependence on history, and so on. But the understanding that has been developed in modern labora-
we also recognize that we need to be able to carry out simple tory testing can be applied to useful effect by practitioners. He
calculations which will be approximate but nevertheless useful has reduced the complexity to a small number of parameters,
as an aid to design. In fact, however sophisticated we make the and it was interesting to see how soils could be compared with
calculations, they are always only rather remote approximations other common materials using these parameters. He has con-
to real behaviour, which contains too many complexities for centrated on a single problemthat of settlement of a shallow
accurate denition and analysis. foundationand has started from the two parameters which
My own rst knowledge of John Atkinson, some time before are easiest to measure: strength at large strains and stiffness at
I met him, was in studying his PhD thesis. This was on the very small strains. Then he has shown how settlements may be
stiffness properties of London Clay measured in triaxial tests assessed for the full range of loading from zero to failure, on
and it provided a lot of insights into the anisotropic behaviour the basis of the stiffnessstrain relationship obtained in labora-
of the material. I noted, in particular, the very useful, and tory tests. This is a problem that has always confronted
honest, statement that he had not been able to resolve strains of designers of foundations: we are better at estimating ultimate
less than 01% (I think it was). Somehow that statement con- failure, which is generally a remote and almost irrelevant
tained the seeds, and the suggestion, that the behaviour of the possibility, than we are at assessing settlement in service,
clay at very small strains might be different from what had which affects all structures, sometimes causing signicant
conventionally been measured. And John has gone on from that, damage. The need to assess displacements is rightly empha-
over the years, to have a leading role in the business of measur- sized yet again by modern codes of practice, but the means to
ing very small strainswhich are really strains of the order do it are often not available.
experienced by the ground in many practical situations. John's principal aim, I know, was that engineers in practice
In the Lecture he has shown us the results of developments should feel that they understand better the signicance of non-
with small strain gauges and with bender elementsshear wave linear behaviour in soils, and that a simple method has been
testingto measure the smallest range of strains. He has re- shown to them, which they can use to make useful calculations,
minded us that interpreting the results of laboratory testing is in line with recent developments in understanding. I think you
not always simple, and poor interpretation can make the soil will agree that John has achieved what he set out to doand
behaviour seem more complicated and less predictable than it his contribution will be tried, tested, used (and he would expect
really is. A clear understanding of the details of test procedure it will be challenged and improved) by engineers in practice.
and an intelligent inspection of the laboratory specimens during I am sure you will wish to join me in thanking John for a
and after test are vital if pitfalls are to be avoided. John and his most interesting and thought-provoking lecture, one which can
team at City University routinely bring these skills to their test out in everyday practice.
work, and pass them on to the students.

Das könnte Ihnen auch gefallen