You are on page 1of 45

Earth-Science Reviews - Elsevier Publishing Company - Printed in The Netherlands

T H E G E O T E C H N I C A L PROPERTIES OF SOILS

I. ALPAN

Faculty o f Civil Engineering, Israel Institute o f Technology, Haifa (Israel)

It is easy to distinguish those who argue from


fact and those who argue from notions...
The principles of every scienceare derived from
experience.
ARISTOTLE

The sciences, blown up by facts, need a


perpetual slimming diet.
RAYMONDQUENEAU

SUMMARY

The solution of problems in soil engineering requires a detailed knowledge


of the mechanical properties of soils which are, perhaps, among the most complex
materials to be studied from this point of view. The present paper endeavours
to present a reasonably comprehensive account of the relations governing the
response of soils to applied forces.
An introductory presentation of the aspects involved and their place within
the general framework of the study of material properties is followed by a discus-
sion of relevant methods used in describing and classifying soils. A separate
section treats the important subject of soil water and the factors influencing its
movement through the channel network of the soil skeleton.
The greater part of the paper is devoted to considerations regarding the
fundamental stress-strain-time relationships. The important principle of effective
stress is presented, followed by a discussion of stress-deformation relationships.
The process of consolidation, i.e., the time-dependent dissipation of pore-water
pressure after loading, is treated in some detail. The applicability, within a limited
range, of linear stress-strain relations to soils is discussed followed by an account
of failure criteria and shearing resistance with special emphasis on the prevailing
drainage conditions. The influence of the strain rate on strength forms the subject
of the last section of the treatment of stress-strain-time relations.
In conclusion, several special topics are presented: a discussion of the pres-
sures and volume changes in expansive clays, the behaviour associated with
thixotropy and sensitivity and, lastly, the response of soils to dynamic forces.

Earth-Sci. Rev., 6 (1970) 5-49


6 I. ALPAN

INTRODUCTION

As c o m m o n l y accepted, soil mechanics is a branch o f civil engineering dealing


with problems in which soils and their relevant properties are of primary concern.
The nature and variety o f these problems are well illustrated by the classification
proposed by COOLING (1945) and shown in Table I.

TABLE I
E N G I N E E R I N G SOIL P R O B L E M S 1

Stability problems Deformation problems

Stability of slopes (slips in cuttings; Settlement of buildings (with and without


embankments, hillsides, river banks, sea piles) and structures of all kinds
coasts, etc.);
Earth pressure on retaining walls, quay Deformation of fills and dams
walls, sheet-piling etc.;
Design of earth dams and seepage Distribution of pressure on walls
Excavations and pressures on timbering and pressure on tunnels, conduits, sewers, etc.
bracing;
Bearing capacity of footings, piles, Cumulative deformations under repeated
subgrades for roads and airfields; stresses, e.g., road slabs

1 After COOLING(1945).

In spite of the wide range o f problems listed in the table, it appears that the
soil properties involved in their solution are essentially related to the following
two behaviour patterns: (1) stress-strain-time relationships; (2) permeability to
fluid transport. As a matter of fact, only the first of the two qualifies as a material
property in the rigorous sense o f the definition given by ROSENTHAL (1964) in
that it concerns the response o f the material "soil" to the stimulus of imposed
forces.
The systematic description o f the behaviour patterns indicated above con-
stitutes, then, the study of the mechanical properties o f soils, and the considerable
complexity o f the subject certainly warrants its separate treatment as one major
subdivision o f soil mechanics. The use o f the results o f this study in the analysis
of models o f engineering problems would properly constitute another major
subdivision: the statics and dynamics of soil masses.

METHODOLOGICALOBSERVATIONS

It can be stated with some confidence that the systematic investigation of


the mechanical properties o f soils has been proceeding essentially on two levels:
(1) On the "structural", or physical, level the origin of their mechanical

Earth-Sci. Rev., 6 (1970) 549


THE GEOTECHNICAL PROPERTIES OF SOILS 7

properties is sought in the nature and arrangement of the basic constituents of soils.
The logical culmination of this approach would be the successful formulation of
a "kinetic theory of particulate media" analogous to the statistical theory of
gases. We are, as yet, very far from such a desirable state of affairs.
(2) On the "phenomenological" level, adopted in the vast majority of
investigations, soils are viewed as continua whose deformational response is
studied in large-scale tests, easier to perform but necessarily of limited range and
validity. The great variety of soil types leads here to an ever increasing volume of
experimental work.
All the same, the tendency has been to discover unifying principles, to achieve
economy of thought leading to economy of experiment, hence towards the simpli-
fication of phenomenological investigations (FREUDENTHAL, 1950).
At this stage, a survey of the mechnical properties of soils must, therefore,
strike a compromise: while presenting, whenever pertinent, related physical
aspects, it will mostly have to discuss empirical relations keeping irrelevant
particulars at a minimum and emphasizing, if possible, unifying trends. Thus it
will prove necessary, quite early in our discussion, to treat sands and clays separately
but there would be scarcely any profit in dwelling on the results of curve fitting
which, alas, forms the subject of so much of the published literature.

DESCRIPTION AND CLASSIFICATION

Soils, as observed before, are "particulate media", i.e., substances having a


skeleton of easily separable particles which enclose interconnected and irregularly
shaped voids. The void space may be wholly or partly filled with a liquid, generally
water containing salts in solution.
The so-called "volume-weight relationships" serve to establish a descriptive
vocabulary in soil mechanics: they are based on a schematic representation of soil
in which its three components are "lumped" as shown in Fig. 1, which also includes
the customary definitions. A basic characteristic of the soil skeleton is the size
and shape of its constituent particles: in sands they are bulky and relatively
large (repr. size 0.5 mm) in clays they are plate - - or needle - - shaped and quite
small (colloidal size) . Both size and shape influence the important soil property
of specific surface which may range from several hundred square metres per
gram for certain clays to a fraction of a square metre per gram for sands. Grain
size is expressed in terms of an equivalent diameter which, for sizes larger than
74 It (B.S. or A.S.T.M. Sieve no. 200), is determined by sieving and below that
size by elutriation methods based on Stokes' Law. The grain size distribution of
a soil is usually represented by a cumulative curve such as shown in Fig.2, from
which various distribution parameters may be determined. One of the most widely
used is the "uniformity coefficient" defined as Cu = d6o/dlo.
Since the soil water considerably influences the behaviour of clays, a number

Earth-Sol. Rev., 6 (1970) 549


8 I. ALPAN

/-"olurne~ ~le~ahls

, I
LIn/~ A r e a

~,q,rnbol and
Term Un /"t
-Oef'/n/(/on

f. Poro$i~/

e. k/O/C/Pat~o e= ~/~
.~ Perce,~-/c,~,e Rzr l/olume a= v o / v c*/*~

~- ~ / ~ ~*/,)

a ffeqree o f ~ u r a l ~ O n <*/o)

6 l~lal Um/( A/eiq/~) ~= wlv qlcm~ 14~/cu.~

z D,-~ un,'t We~qh~ ~ ~/v q/cms , /6s.lcu.f~

Fig.l. The three soil constituents.

1oo

~Jo/!groded- 1o~I un/~orrn/t

1900rILl 4to.feel-h/g]~ Un/iCorml/~/

2/ScOnf/nuoua - frc~c[/on rn/ee/;~Sz

Fig.2. Typical grain size distributions. Uniformity coefficient: Cu = dno/dlo >t 1.

o f simple tests have been devised in order to assess the susceptibility o f such
soils to moisture changes. These tests serve to establish the moisture range related
to plasticity and shrinkage and the results are given as moisture contents termed
"consistency limits". In almost universal usage are the following (details e.g.,
LAMBE, 1951): The liquid limit, wj, the plastic limit, Wp, and the shrinkage limit,
ws, as well as indices derived from these, for example: the plastic index, Ip = w I - wp

Earth-Sci. Rev., 6 (1970) 5 ~ 9


THE GEOTECHNICALPROPERTIESOF SOILS 9

and the consistency index, I c = (w~ - w ) / I p where w is the natural moisture content.
Others of these so-called "index properties" will be presented as the need arises.
In our context, a soil classification system, in order to be useful, would be
directed towards connecting the mechanical properties of typical soils with simply,
rapidly and economically determined index properties such as those presented
above. Indeed, as can be gathered from an excellent review by L1u (1967), granul-
ometry, limits and indices form the basis of most systems in use. The "plasticity
chart" shown in Fig.3 and due to CaSAORANDE (1948) is an illustration of a
qualitative grouping of fine-grained soils in a I p - w~ coordinate system.

L/quid Z/rrl/z" ~ (%)


7"0
tO ~0 JO 40 60 dlo ~ ~O 90
I I I ] l I I I I

O0

OrOorl/~ Cloy.5 oK . /
5.0
~/, P.'os/.~//..z .~o~V')/

.0
I ftTor~Tort/c
.f
3O
Ic/o~,8 of
Mcr'/~'m
Oi'qa/r,c C,~qs Plos~,'e~/(l
o f Zo~."
~ c Nl,/5of"
0
Co/zes/~/e~s j

0 so;v~ , , / 1 I I r I 1
I
~orqc,.-tzc $,//s Lf n o r q o n : c ~ , h ' ~ of/'leo"/u,..r~
Of LOal C o m p r ~
-~,~,./,.~.
ConTprosBl bil/'~/ Or~dOr~nie
Fig.3. The plasticity chart. (After CASAGRANDE,1948.)

In addition, it is of obvious usefulness to investigate possible correlations


between basic properties of soils and their more easily determined index properties.
By suitable elimination of irrelevant variables or parameters certain relationships
between soil properties may become apparent suggesting improved and more
rational testing procedures.

SOIL WATER AND ITS MOVEMENT

The study of soil water is of considerable importance in soil mechanics, both


as a factor influencing the mechanical properties of soils and regarding its flow
through the available pore space. As an illustration, Fig.4 shows the influence of
water content on the response of a plastic clay to shearing stresses.
It should be emphasized that water as encountered in soils contains, in
general, electrolytes in solution, a fact which adds to the complexities of analysing
its behaviour (e.g., DORSEY, 1940; FORSL1ND, 1952; LOW, 1961; ROSENQUIST, 1961).
In presenting a classification of soil water, it should be pointed out that

Earth-Sci. Rev., 6 (1970) 5-49


lO 1. A L P A N

20

fO - - . --
i
d~
7~
~06 -- -
6

02 2
30 J5 .40 .~5 Jo
LdoZeP C o n ~ e ~ Z - L,/ f ~ )

Fig.4. The influence of water content on the failure conditions of a plastic clay. (After
NORTON, 1952.)

numerous attempts have been made in this direction, each designed according to
a particular point of view. The following, rather broad, system appears to be quite
generally accepted (ROAD RESEARCHLABORATORY, 1952, ch. 16; BEAR et al., 1968)
and is illustrated in Fig.5.

So//uoter

We~#u~ter
6round GroL//~og/onol I
i/crier
[
~o~r

~,e.,d ~po.r
/~wose
Fig.5. Soil water classification.

The water filling the soil voids below the water table (where, by definition,
the water is at atmospheric pressure) is called "ground water", whereas the water
percolating through the soil under the action of gravity is termed "gravitational
water". Some of the water, instead of draining, may be retained in the intricate
channel network of the voids, constituting what is termed "held water", part of
which may be present as vapour.
A more detailed picture regarding the nature of soil water is presented in
Fig.6, following MICHAELS (1952) and BEAR et al. (1968).
The "pore water" is normal liquid water displaceable by ordinary hydro-
dynamic means provided there are no excessive capillary forces present. The

Earth-Sci. Rev., 6 (1970) 5-49


THE GEOTECHNICALPROPERTIES OF SOILS 11

,.. /---/

/ , /g# , k -. .- .- . . b_'%, )
i x _AY.z--. . . . - : - :-
(' ~'~./ t .... _t . . __
--__~t-L . . . .__. __ \j
Rdsorbgd 6o~al/On ,Oor~ &laar
LJater l ~/alar ~. , . ~ I
I I ~ap///crr 6rbund ~Mob/h LJa~er
Ualar [daTer | (~r~e)
Held LJoTer
Fig.6. Types of soil water. (After MICHAELS,1952.)

"solvation water" surrounds the soil particles in relatively thin layers (20-200
molecules deep) and is held by polar, electrostatic or hydration forces near the
particle surfaces. This type of water is considerably denser and more viscous than
ordinary liquid water but is, nevertheless, still mobile. The "adsorbed water"
covers the particle surfaces in a very thin layer (up to 100 A) held by extremely
large forces and it may be considered as essentially immovable by hydrodynamic
mechanisms. "Structural water", finally, should not be considered as water at all,
being in fact an integral part of the crystal lattice of the particles, and its removal
is associated with the structural breakdown of the constituent minerals.
It is evident that the basic approach in the classification system discussed
above is connected with the concept of the retaining forces acting on the soil
water and governing its mobility (LeKow, 1958, table I, p.2, 3). Instead of de-
scribing the force field by a vector space function (V) it has proved more convenient
to use a scalar potential function, ~9, where V = grad ~, as proposed by BUCKING-
HAM (1907) with respect to capillary retaining forces.
The soil moisture potential is defined as the minimum energy per gram of
water required to transport an infinitisimal test body of water from a specified
reference location (usually a free flat water surface at atmospheric pressure) to
any point within the liquid of the soil-water system at rest (BOLT and MILLER,
1958). It appears logical that, following further research into the problems of
soil-water energetics, the concept of soil moisture potential should come to include,
in addition to the gravity and pressure components, also those related to osmotic,
electrical and thermal effects (RusSELL, 1942; HABIB and SOEIRO, 1957; BOLT and
MILLER, 1958).

Earth-Sci. Rev., 6 (1970) 5 4 9


12 I. ALPAN

The temporary fixation of water in soil may, in general, be attributed to the


adsorption of water molecules on the particle surface and to surface tension
forces. Hence, the removal of water from the soil pores obviously requires a
certain amount of work and, for the force used to effectuate it, the term "suction"
was proposed by SCHOFIZLD(1935).
Suction, as defined, is the tension (pressure below atmospheric) to be applied
to the pore water in order to initiate its removal at a certain moisture content,
the porous medium being free from any external loads. As a convenient scale
for suction, the common logarithm of an equivalent negative water head in cm is
used, designated by the symbol " p F " . Thus a suction of 1 kg/cm 2 has a pF value
of 3. A pF-curve for a clay, exhibiting typical hysteresis behaviour, is shown in
Fig.7 (COLEMANand MARSH,1961).

70

1,,, ao

U owen ~ ~,,~..-~

fO

0
o io 20 30 40 ~0

Mo/~turo contenZ - (O/o)

Fig.7. Suction curves for a clay. (After COLEMANand MARSH, 1961.)

Turning now to the problem of moisture transfer in soils, we shall limit the
discussion to mass flow under the influence of a hydraulic gradient, this being
the flow mechanism most often of practical interest. Under conditions of full
saturation, the specific discharge q is related to the driving hydraulic gradient J,
i.e., loss of head per unit distance of flow, by the famous empirical relation of
Darcy:

q = kJ (1)

where the proportionality constant k is called the "coefficient of permeability".


For saturated laminar flow, the coefficient k is, in general, dependent upon the
following properties of the percolating fluid and the percolated medium:
(1) fluid properties: unit weight, dynamic viscosity;
(2) medium properties: grain size, void ratio, pore geometry.
A rational relationship of this sort would be the following (TAYLOR, 1948,
ch. 6):

Earth-Sci. Rev., 6 (1970) 5~19


THE GEOTECHNICAL PROPERTIES OF SOILS 13

k = D2 Yw e3 C (2)
pl+e

where: Ds = representative average grain size; 7w = unit weight of fluid; p =


dynamic viscosity of the fluid; e = void ratio ( = n/l-n; n = porosity); C =
shape factor related to pore geometry.
In unsaturated flow, Darcy's rule may be assumed to hold but, obviously,
the coefficient of permeability is influenced by an additional factor, the degree
of saturation, S. If k o denotes the saturated Darcy coefficient and So a lower
limit of the degree of saturation (corresponding to an "irreducible" water content),
the following relationship has been proposed (BEAR et al., 1968, ch. 9):

k _(S-Sol 3
ko - So/ (3)

It is seen that a relatively slight decrease from full saturation considerably reduces
the specific discharge to be expected under the same gradient.
Relations analogous to that for flow induced by a hydraulic gradient appear
to be valid for flow under the influence of thermal or electrical gradients. Thus,
from the data published by HABIB and SOEIRO (1957), the specific thermo-osmotic
discharge (at full saturation) may be expressed as:

dO
qo = k ~ ( 0 m ) - - (4)
dx

where: k = Darcy coefficient; ~(0m) = a constant dependent upon the mean


temperature, 0m; dO/dx = temperature gradient.
For electro-osmotic flow, L, CASAGRANDE(1952) presents the relation:

dE
q~ = k e - - - (5)
dx

where E = electrical potential.


The permeability coefficient k e is shown to depend upon soil porosity, the
dielectric constant, the viscosity of the liquid and the electro-kinetic potential
related to the interaction between charged soil particles and the ions of the soil
water (SHAw, 1966).

STRESS-STRAIN-TIME RELATIONSHIPS

Stated in its simplest terms, the problem under discussion in this section is
to describe the deformational response of soils to the stimulus of applied stresses.

Earth-Sci. Rev., 6 (1970) 5-49


14 I. ALPAN

The basis of such an enquiry must be the physical behaviour of the material which
determines the choice of an adequate mathematical model aimed at a quantitative
and concise description of this behaviour.
From the start, the analysis is complicated by the fact that our material is
an aggregate of individual interacting elements (particles, water and air) and
that, consequently, its behaviour may be expected to be governed by constitutive
properties as essentially determined by the predominating group pattern (FREUDEN-
a'HAL, 1950). Furthermore, the dominant role of the operating interaction deter-
mines a pronounced time-dependence of the deformation response of soils, aptly
described by BERYATZn(1947) as "pseudo-solid bodies", to which the conventional
methods of strength of materials can be applied only as a first and necessarily
crude approximation.
As a consequence of this state of affairs, the problem has been attacked on
two levels: one characterized by an endeavour to deduce the relevant properties
from a particulate model, the other by a phenomenological approach treating soils
as a continuum (e.g., TAN, 1957; DERESIEWICZ, 1958; Kt~ZDI, 1966). Although
conceptually very useful, the first approach has been, so far, only moderately
successful mainly, in my opinion, because of the exceptional difficulty in formulating
the statistical relationships required by this model. On the other hand, the con-
tinuum hypothesis, ignoring the particulate nature of real soils, conveniently
postulates the existence of a material whose dynamical and kinematical variables
are continuous functions of space.
The preceding generalities, although necessarily brief, may have illustrated
the difficulties of presenting a systematic picture of the deformation properties of
soils. Accordingly, no more will be attempted in the following than to discuss the
response of the two basic soil types--sands and clays--to applied stresses, in the
belief that the discussion of composite soil types would add little beyond empirical
information, however useful in engineering practice.

Effective stress
If a total stress tensor S---customarily represented by the three principal
stresses al, a2 and ff3--is applied to a soil specimen, it is in part transmitted to
the soil skeleton and in part to the pore fluid. The stress component transmitted
to the skeleton is taken to be effective in determining the deformational response
of the soil, hence the heuristic principle of effective stress which has become one
of the most important and, at times, controversial concepts in modern soil me-
chanics (SKEMPTON,1960; JENN1NGSand BURLAND,1962). This principle involves,
it would seem, both a definition and a physical law"
(1) by definition, effective stresses govern the strength and deformation
characteristics of soils;
(2) the physical law establishes, with all the accepted reservations, the re-

Earth-Sci. Rev., 6 (1970) 549


THE GEOTECHNICAL PROPERTIES OF SOILS 15

lation between the effective stress, the total stress and the pore fluid pressure
existing, at a given instant, in a soil.
Thus, for saturated soils:

a' = o - u (6)

where: a' = effective stress; cr = total stress; u = pore water pressure; (pressures
are taken as positive in soil mechanics, whereas the resulting volume decrease is
taken as negative).
For partly saturated soils, the following relationships has been proposed by
BISHOP (e.g., 1959):

a' = tr - u a + Z(ua - uw) (7)


where Ua = pore air pressure; Uw = pore water pressure; Z ---- parameter Varying
between 0 (dry soil) and 1 (saturated soil). The validity of the law is, of course,
subject to experimental confirmation.
Now, as far as saturated soils are concerned, RENDULIC (1937) has offered
excellent p r o o f by correctly predicting the pore water pressures set up in undrained
shear. As for unsaturated soils, the p r o o f is based on the character of the governing
equation: if, at a particular instant during shear, X is considered constant, then:

do' = d(a - u,) + z d ( u a - Uw) (8)

at this moment, even if the individual values of a, u~ and Uw are suddenly changed.
Experiments by Bishop and Donald (quoted by SKEMPTON, 1961), in which
the bracketed expressions were each kept constant, showed this to be the case.
The above reasoning is not invalidated by the complex character of z - - i t appears
in eq.8 quite legitimately as an empirical operator. From physical considerations

o8

/
D2

0 20 ,4O 60 8o Ioo
De3r~e of S~tvr~tZo~ (*~)

Fig.8. The parameter Z vs. degree of saturation. (After BISHOPand BLIGHT,1963.)

Earth-Sci. Rev., 5 (1970) 5-49


16 I. ALPAN

(ALPAN, 1965a), the approximate equality of the parameter Z with the degree of
saturation, S, may be deduced. Empirical evidence, such as presented in Fig.8
(BISHOP and BLIGHT, 1963) shows Z as a somewhat more complicated, though still
monotonous, function of S.
The preceding remarks, though brief, should suffice in emphasizing the fact
that any fundamental inquiry into the mechanical properties of soils requires a
knowledge of the pore fluid pressure changes accompanying the variations in total
stresses applied to a sample.

Stress and deformation


Now, in principle, a soil sample can be tested as a closed or as an open system.
In the first case, the sample is isolated from its surroundings by an impervious
membrane which prevents any escape of pore fluid and through which the total
stresses are applied; in the second case, the pore fluid is allowed to drain. In both
cases it is possible, though not always necessary, to measure the variations in pore
fluid pressure, as discussed in great detail in connection with testing techniques
(e.g., BISHOP and HENKEL, 1962; SOWERS, 1964).
At present, the most versatile equipment in general use in soil testing is the
triaxial apparatus in which cylindrical specimens are acted on by a triplet of
principal stresses of which at least the two radial ones are always equal, i.e.,
either 0.1 > tr2 --- G3 or 0.1 < 02 = 0"3" Pore pressure measurements are possible
either at equilibrium, when the soil is tested as a closed system, or in following the
pressure variation at a particular point (usually one end of the specimen) if drainage
is allowed.
In the oedometer, uniaxial deformation of an open system is measured under
conditions of zero lateral strain. In general, no pore pressures are measured, as
the tests are taken to last sufficiently long for any pore pressure to dissipate as the
water drains from the sample.
The direct shear box, again, is used to apply to an open system a combination
of vertical and tangential stresses on a predetermined plane along which the soil

TABLE II
T H E T E S T I N G OF SOILS

Apparatus Open system Closed system Remarks

Triaxial consolidation shear controlled principal stresses any K


apparatus compression compression (or2 = or3 = K~rl)p.w.p, can bemeasured
shearing strength
Oedometer controlled major principal stress, (~rl =
(consolidometer) consolidation Ko ~a)
compression
Direct shear box shearing strength controlled Crn(principal stresses vary in
magnitude and direction)

Earth-Sci. Rev., 6 (1970) 5~19


THE GEOTECHNICALPROPERTIESOF SOILS 17

is brought to failure by sliding. Here, too, the pore pressures are unknown and
must be inferred, among other things, from the time-rate of shear and the soil
permeability (GIBSON and HENKEL, 1954). Table II summarises the discussion
above while some of the concepts shown will be dealt with further on.
It will be convenient to describe changes in stress as "stress paths", that is
lines in an orthogonal stress space whose axes represent the three principal stresses.
In practically all tests, the relevant stress paths are restricted to a bisector plane,
n, such as shown in Fig.9.

\\\\
[iE' c~ ~/r@SS 8pQce

//?'~IP

/ \'/ I /. . u.

/~ B/seclor Pl~ne17

/'~\[pur. d..,a~/orZc/o~d,'~,g
/ x,

,.7
v 1 I
Fig.9. Stress space and stress paths.

Consider a soil sample undergoing all-round compression as a closed system,


i.e., being exposed to total stress changes as represented by the stress path OA in
Fig.9b. The ratio of the induced pore pressure change, Au, to the applied hydro-
static pressure, Ap = ActI = Aa 2 = Ao-3, turns out to be (SKEMPTON, 1954):

Earth-Sci. Rev., 6 (1970) 5~19


18 1. ALPAN

Au 1
B . . . . . . . (9)
ap .o cf + (1 1
C C

where: n o = the initial soil porosity; C = the compressibility of the pore fluid;
Cs = the compressibility of the soil particles; C = the compressibility of the soft
skeleton (in terms of effective stresses).
For fully saturated soils, the compressibility ratios in eq.9 are very small compared
with unity, hence:

(B)s=loo% ~ 1 (S = degree of saturation) (10)

In partly saturated soils, the ratio Cr/Cc cannot be neglected and the value of the
pore pressure coefficient B becomes appreciably smaller than 1.
For this case, the average compressibility of the pore fluid becomes (BISHOP
and ELDIN, 1950):

C r-= (1 - S + S H ) / p (1l)
where: H = Henry's coefficient of solubility ( ~ 0.02 for air in water); p =
absolute equilibrium pore air pressure. It follows from eq.6, 9 and 10, that the
change in effective stress in all-round, undrained compression is:

Ap' = (1 - B ) A p (12)

and approaches zero for fully saturated soils.


The relative volume change
AV
- Cc(1 - B) Ap (13)
Vo
similarly vanishes at full saturation, quite in accordance with the principle of
effective stress.
I f drainage is allowed to take place, Au equals zero at equilibrium and,
therefore, Ap' = Ap.
It is evident that any stress triplet, falling outside the "hydrostatic" stress
axis, (where AGI = Aa2 = Ao-3), implies the application of shear stresses having a
maximum value of A%,ax = (A1--Aa3) for Act1 > Aa 3 ( = Acz). Let the stress
triplet in question be represented by the point P in Fig.9b; it can obviously be
reached by a variety of stress paths of which the following are shown as especially
significant:
(1) OA-AP." all-round pressure followed by the application of shear stress
with AO" 2 = Air 3 held constant.
(2) OP: the stress ratio, K = Atr3/Aa I = constant.
(3) O B - B P : all-round pressure followed by a stress change during which the

Earth-ScL Rev., 6 (1970) 5M-9


THE GEOTECHNICALPROPERTIESOF SOILS 19

first stress invariant, AI 1 = Aa~ + Aa 2 + Aa 3 is held constant. B P is frequently


termed the path of "pure deviatoric loading" (KLAUSNER, 1964).
The remarks above apply, naturally, to total as well as effective stress paths
depending upon the conditions investigated.
The basic object of stress-strain investigations in soil mechanics appears,
thus, to be the study of effective stress paths and the concomitant changes of
volume, shape or both.
In the general case of triaxial stress application (Aax > A~r2 : Aa3) we may
write (SKEMPTON, 1954):

Au = B[Aa 3 + A(Aa~ - Aa3) ] (14)


and, based on concepts by SKEMPTON and BISHOP (1954):

AV C~Aa3/1 _ B + (Sd _ BA) [ Aal 1]~ (15)


Vo [ Aa3 ]
where A is a pore pressure coefficient and S d is defined as a structural parameter.
At full saturation with drainage prevented, B = 1 and, since AV/V o = O,
we find that:

Sd = A (16)
For all-round compression, with Aal = AO-z = Ao-3 = Ap, eq.14 and 15 become,
of course, identical with eq.9 and 13. It may be noted, in passing, that eq.16
appears to confirm the so-called "American hypothesis" regardi.ng the shear
strength of normally consolidated clays (SKEMPTONand BI sHoP, 1954, p.473).

Consolidation
As mentioned before, the process of drainage in a fully saturated sample is
accompanied by the dissipation of the pore water pressure, (Au ~ 0 ) , and by
volume changes resulting from the simultaneous changes in effective stress. It is
obvious that the rate of this process must depend, in the first place, upon the soil
permeability. In clays, the process, termed consolidation, is in most cases a very
slow one and has, for this very reason, formed the subject of extensive studies
over forty years since the classical investigations of TERZAGHI (1923, 1924) and,
later, of BlOT (1941) and TAYLOR (1942).
The conventional analysis of saturated consolidation, in briefest outline,
is as follows (cf., GmSON and LUMm 1953):
Let the void ratio e, be a function of the applied effective stress, i.e.:

e = e(a') = e(a -u) (17)

then:

Earth-Sci. Rev., 6 (1970) 549


20 I. ALPAN

de = __3e d a ' = __3e (da - du) (18)


Oa' Oa'

If, as usual, the applied total stress remains constant during the process, and the
stress interval is sufficiently small:

0e
- -- = av = the constant "coefficient of compressibility". (19)
0t7'

However, for larger stress intervals applied to "normally-consolidated" clays, i.e.,


clays being consolidated under a pressure never exceeded in their past, the fol-
lowing relationship holds:

e = eo - Cc log (a'/a'o) (20)


where Cc = constant "compression index".
It follows that, for this case:

0e
- 2.3 (Cda'). (21)
Off'

The time rate of change of the void ratio can be written as:

0e 0u
- av- (22)
Ot Ot
Assuming the validity of Darcy's rule, the application of the law of mass con-
servation yields:
3
0 2u av 0U
1 k~ - (23)
Yw ~ 1 + e at
i=1
and, for a homogeneous and isotropic soil:

CV2/,/- 0U (24)
Ot
where the "coefficient of consolidation":

k(1 + e)
c - (25)
~w " av
may be assumed as sensibly independent of the applied effective stress.
The analysis, outlined above, while forming the basis of practically all work
on the consolidation problem, fails to account for the relative movement of the
soil skeleton and the extruded water during the process. An attempt to include
this aspect was presented by ZASLAVSKY(1964), ultimately leading to the following
expression:

Earth-Sci. Rev., 6 (1970) 5 4 9


THE GEOTECHNICALPROPERTIESOF SOILS 21

Vk.V (y~
u + z ) = ~ - ~3 [ln(1 + e)] + qs. V[ln(1 + e)] (26)

where qs = average velocity vector of the soil particles relative to a fixed co-
ordinate system. Another approach, claiming greater generality, has been
proposed by MIKASA (1965), who formulates the consolidation process in terms of
a logarithmic void ratio function as follows (modified here to correspond to eq.
24):

c" V2e - & (27)


at
where:

e----In .1. .+. e o (28)


l+e
However, as already indicated, eq. 24 represents the generally accepted formulation
of the theory of consolidation and its solutions, for various boundary conditions,
form the basis of settlement predictions in engineering practice. Of these solutions
we shall briefly consider the axial-symmetrical case, i.e., that of vertical and
radial drainage of water from a saturated cylindrical soil volume (CARRILLO,
1942; GIBSON and HENKEL, 1954).
Let Uo be the initial, uniformly distributed, pore water pressure in a saturated
cylindrical clay sample prior to the start of drainage and u(t) the average pore
pressure in the sample at a time t after drainage has started. The average degree
of consolidation is then defined as:

U
U = 1 -- (29)
Uo
If drainage takes place both in the vertical and radial direction, the average degree
of consolidation is shown to be:

U -- 1 -- (1 -- Uz)(1 - Ur) (30)


where:

Uz = 1 - -uz
- (31)
UO

and:

U r -- 1 ~/r (32)
Uo

Here, Uz is the solution of the one-dimensional consolidation equation:

Earth-Sci. Rev., 6 (1970) 5-49


22 I. ALPAN

92/,I 0u
cz ~z ~ = -Ot (33)

and Ur that of the radial consolidation equation:

F02. + )_2u
Cr[~?2 F OF
] = 5t
0u (34)

For a cylindrical specimen of height h and radius R, with only vertical drainage
through both end surfaces, the solution of eq.33 introduced in eq.31 furnishes:

8 l 7~2 Cz
Uz = 1 - - - ; exp - ( 1 + 2n) 2 - t (35)
4
n=0

with H = h/2, the longest drainage path; whereas for radial drainage, the solution
of eq.34 introduced in eq.32 yields:

1 2 Cr
Ur = 1 - 4 ~-exp - o, ~ t (36)
On
n=l

where o), is the nth root of the zero order Bessel function, Jo-
In most cases, due to stratification, cr/c z = ~ > 1 (e.g. ABOSHt and MONDEN,

tO
1 I
1 U=SO %
%

-i i

.o, i[
0 / ? ,.7 4 6 5 7 S 9 tO

~. = Cr/C z

Fig.10. The influence of anistropy on the speed of consolidation.

1960, 1961; McKINLAY, 1961) and the influence of this condition is illustrated in
Fig.10 for 5 0 ~ average consolidation, where ~ = 0 represents one-dimensional
(vertical) consolidation, the time for this case being taken as reference.

Earth-Sci. Rev., 6 (1970) 5 4 9


THE GEOTECHN1CAL PROPERTIES OF SOILS 23

Elastic theory concepts


Reverting to our general topic of stress-strain relations, one would expect
to find a considerable amount of research concerned with attempts to express these
relations in terms of conventional parameters such as Young's modulus, Poisson's
ratio etc., mainly as a result of an unfortunate dilemma. For, while everybody is
aware of the non-linear characteristics of soils, the analytical tools used in the
majority of deformation problems in soil engineering are based on results from
linear continuum mechanics e.g., the problem of the stress distribution in soil
masses. Still, if unwarranted extrapolations are avoided, very useful results are
available.
As a typical and interesting example, the work of JAKOBSON (1957) on sands
can be cited. Fig. 11, based on his results, shows the variation of E and v as functions
of the effective stress ratio K = a3/a~ applied in the triaxial apparatus. Similarly,
the function K/(I+K) is shown which, in elastic theory, corresponds to zero
lateral strain. The intersections of this function with the v-lines of the tested
sands would then represent the so called Ko condition (e.g., ALPAN, 1967a) and
Q7

06 .- -- .

.~
O.~ - - 7~OO "--

"~ 0.4 f _ 2aO0

"

O.t _ / 7 ___ 500

J
o
0 02 0.4 06 O0 fO

Fig.l 1. Elastic 'constants' for sands. (After JaKOBSON,1957.)

it appears that the value of Ko, as determined for the specific testing conditions
imposed, is greater in the looser sand, contrary to results obtained in conventional
oedometer tests. On the other hand, the evidence of Fig.11 indicates an increase
of Poisson's ratio with density, in agreement with a tentative interpretation of
certain dynamic tests on sands (ALPAN, 1967b).

Earth-Sci. Rev., 6 (1970) 5-49


24 i. ALPAN

The initial portion of a typical stress-strain curve is, in general, sufficiently


linear to permit the determination of a reasonably constant modulus for this range,
usually termed "tangent modulus", E~. The variation of this modulus with over-
burden pressure is shown in Fig.12 (cf., T~RZACHI and PECK, 1967, art. 15); also

500O
J II I
" / q 2~/rn~r /

o :
5 tO L~O ,~ 40 5t9

Z~ePfh of ~/erL~JrDCen - m

Fig.12. Tangent modulus vs. overburden. (After TERZAOH[and PEck, 1967.)

shown, for comparison purposes, is a curve for a sand at 7 0 ~ relative density


computed from data given by WILSON and SUTTON (1948) exhibiting a similar
trend.
The problem of the linearity range, mentioned above, has been recently
investigated by SOUTO SILVEIRA(1967) who defines the "linearity percent" as that

6'5

I"

/'7o/~/u.re Co/7/enz" :~/ (O/o)

Fig.13. Linear deformation range of a soil. (After SOLrTOSILVEIRA,1967.)

Earth-Sci. Rev., 6 (1970) 5--49


THE GEOTECHNICAL PROPERTIES OF SOILS 25

fraction of the failure stress which constitutes the upper limit of that range. As
evident in Fig.13, this parameter is seen to decrease with increasing moisture
content and the all-round consolidation pressure in the triaxial cell.

Soil strength
In the preceding paragraph, the failure stress appears as a convenient
reference stress and it is appropriate, at this stage, to continue our discussion of
stress-strain relationships with a somewhat detailed treatment of the problem of
strength and failure in soils.
Soils, strained in shear beyond a certain limit of resistance, are considered as
having failed and thus strength may be defined as resistance to excessive shear
deformations.
A typical triaxial compression test curve, corresponding to the stress path
A-P of Fig.9b, is shown in Fig.14 with the customary failure condition at the
maximum principal stress difference.

~-~

too [ "

I , Ic

1 I \1 I
Fig. 14. Triaxial stress-strain curve.

Once the allowable limit of deformation has been stipulated, the state of
failure can be described by a relationship involving stresses and material parameters;
such a relationship constitutes a failure criterion.
For isotropic materials, failure is independent of orientation, hence, ac-
cording to the general mathematical principles of invariance, a scientifically correct
failure criterion must contain the three principal stresses in a cyclic-symmetrical
a r r a n g e m e n t (BRINCH-HANSEN and LUNDGREN, 1960, p.39; IRMAY, 1968). Failure
criteria for soils would evidently be expressed in terms of effective stresses. Thus,
the generally accepted " M o h r - C o u l o m b " failure criterion may be formulated as
(DRUCKER and PRAGER, 1952):

Earth-Sci. Rev., 6 (1970) 5~-9


26 i. ALPAN

~(Allf - 3Auf) + "V/[ - A/d2]f = k (37)

where:

Allf = 3(A6oct) f = Aa~r + AO'2f "4- Ao-3f (38)


and:

l a i d 2 ] f = _ ~(Aro,)f
3 2 =
-- ~[(AO-lf -- A o 2 f ) 2 -1- (Ao'2f - A o 3 f ) 2 -~- ( A o 3 f -- AO'lf) 2] (39)
and c~ and k are material parameters.
In more conventional notation, the failure condition in triaxial compression,
(AO-l' f > A o ' ; f ~-- /~O'if), reads:

AO"lf = N 0, Ao-3f + 2c'a/N O, (40)


with the "flow value"

1 + sin qS'
N 0, - (41)
1 - sin qS'
where qS' denotes the effective angle of shearing resistance and c' the effective
cohesion intercept. In terms of the effective normal and shearing stresses on the
failure plane (inclined by 0 r to the major principal plane) the failure condition
reads:

A~-f = c' + Ao-'nftan ~b' (42)


where:

Ao-'nf = (1 - sin qT)Aa'lf - c' cos qS' = (1 + sin ~b')Ao'3f + C' COS q~' (43)
The failure conditions expressed by eq.37, 40 and 42 are shown in Fig.15 for a
soil of given strength parameters.
We shall continue in discussing the strength characteristics of sands and
clays separately.
Sands do not exhibit a cohesion intercept, i.e., c' = 0, k = 0. Consequently,
their shearing resistance is completely defined by the p a r a m e t e r q~' which depends,
in general, on the following physical properties (BR1NCH-HANSEN and LUNOGREN,
1960):
(1) G r a i n shape: the change in ~b' between the extremes of very angular and
very r o u n d grains is of the order of 6.
(2) Grain size: for the range sand-gravel a change in $ ' m a y be of the order
of 2 . There appears, however, to be a m a r k e d influence of gradation: thus, uni-
f o r m materials exhibit a much larger change than well-graded ones, namely
of the order of 10 f o r a range of m a x i m u m particle size from 0.2 to 20 mm.
(HENNES, 1952).

Earth-Sci. Rev., 6 (1970) 5-49


THE GEOTECHNICALPROPERTIESOF SOILS 27

3~0

~o .// !

zo ,. No.O,," c/;,~c~

I to
0 /.O lifo ,3".0 4.0 .o.0
i I I
1i role' ~;r ,~m~ I,r-a~ur
Jme
Fig.15. Failure conditions.

(3) Grain size distribution as expressed, for example, by the uniformity


coefficient C, = d 6 o / d x o .
(4) Density of packing, commonly expressed by the void ratio, e.
The influence of the last two properties is illustrated by the empirical curves
shown in Fig.16 which are based on the following relationship found to express

<'

/.0 ~ . . .
] .

o N"kN',

0.4
02 O..r 04 05 Oa 02" 08 o9 Io iI
ZnitlUI VOid rotlO - eo

Fig.16. The influence of void ratio and grading on the shearing resistance of sands.
adequately the variation of shearing resistance with void ratio (CAQUOT and
K~RISEL, 1966; WINTERKORN, 1966):

Earth-Sci. Rev., 6 (1970) 549


28 L ALPAN

e tan ~b' = ks (44)

where the parameter k s is shown to depend primarily on grading. To some extent


ks decreases with increasing effective confining pressure (cf., TAYLOR, 1948, ch. 14:
Tests on Ottawa sand). The interesting fact of an apparent upper limit of strength
in granular materials has been discussed theoretically by IRMAY (1968).
If tested in shear as an open system, sands exhibit volume changes which
depend, in the first instance, on the initial state of packing: dense sands dilate,
loose sands are compressed. This tendency affects the behaviour of saturated
sands when sheared as a closed system: in dense sands the pore water pressure
diminishes while it increases in loose sands.
An additional factor, governing the volume change in sands undergoing
shear, is their effective state of stress and its character. Indeed, recent research
supports the view that, up to a certain level, pure deviatoric loading does not
appear to produce significant volume changes (FRYDMAN, 1968). In any event, it is
reasonable to postulate the existence of a "critical" state of density and confining
pressure in which sands do not exhibit volume changes in shear (for a thorough
discussion see TAYLOR, 1948). The critical void ratio, e , , decreases with increasing
confining pressure, Per, and this relationship determines two regions of potential
volume change in a (e, p)-plane as shown in Fig.17. If, now, a saturated sand is

~CF

ll\ gonlracl/~n Zone

Ca
Uni(orrnltLl oef~/c~n/ cr/Y/c~l c~//~7iny P r ~ u z - e
Fig. 17. Critical void ratio and pressure.

exposed to an application of shearing stresses, quick enough for no drainage to


occur, the pore water pressure will change according to the zone representing the
state of the sand. According for example to eq.40 the factor of safety of the sand
against shear failure can be written as F = N o, (0"3--//)/(0" 1 --//) ~ 1, from which
the change in safety due to a change in pore water pressure is easily derived as

Earth-Sci. Rev., 6 (1970) 549


THE GEOTECHNICAL PROPERTIES OF SOILS 29

having the form d F = - a du, where the constant a > 0 represents the initial
conditions. It follows that failure may occur in the "contraction" zone, provided
the sand in question exists in the corresponding state. That this may not be so is
similarly shown in Fig. 17 in which the relationships of Fig. 16 have been qualitatively
introduced to show that increasing uniformity of grading enhances the danger of
"liquefaction", a term commonly used for the type of failure described above.
The factors influencing the shearing resistance of clays appear to be too
numerous and, in part, not adequately clarified to allow a unified treatment of
the problem. Thus, it is most instructive to study the list of fundamental and
interrelated factors given by TAYLOR (1948, section 15.2) in order to appreciate
the complexities involved. Accordingly, our discussion will have to be limited to
those basic aspects which, while far from exhausting the subject, still a fforda
sufficiently coherent view of it.
In limiting at the outset our treatment to fully saturated clays, a condition
most often encountered, we shall postulate as significant the following three
failure conditions (cf. BRINCH-HANSEN and LUNDGREN, 1960, para.l.42), the
relevant equations being formulated for direct shear:
(/) True failure: determined by the effective stresses at a given void ratio (or
moisture content) regardless of the stress history of the material. Thus in the
relation:

Azf = Cr + Aa',f tan 4~r (45)


the true strength parameters cr and Cr are considered unique functions of the
moisture content at failure. In a rigorous sense, the stress history of a clay should
show the variation with time of the effective stress tensor applied to it from its
formation to the time of testing. In actual fact, however, a comparison only is
made between the maximum effective stress ever experienced by the clay in
question, and the effective stress under which its properties are being investigated:
if both are equal the clay is said to be "normally consolidated", if a larger effective
stress has been experienced in the past, the clay is termed "over-consolidated"
(cf. our discussion of consolidation).
(2) Effective failure: determined by the effective stresses at various moisture
contents. Its formal expression has been given in eq. 42:

Azf = c' + Aa'nf tan '

where the effective strength parameters c' and ' are influenced by stress history.
Obviously, the two failure conditions presented above imply the pore water
pressure at failure to be known.
(3) Apparent failure: determined with respect to the total applied stresses:

AZf : C q- Acrnf tan q5 (46)

Earth-Sci. Rev., 6 (1970) 5-49


30 i. ALPAN

where the strength parameters c and ~b depend, amongst other things, on stress
history and drainage conditions.
Consider a saturated clay, normally consolidated under an effective all-round
consolidation pressure a, and subsequently brought to failure in triaxial shear
with no drainage allowed. Its true cohesion intercept is found to be linearly related
to the consolidation pressure (GmSON, 1953; HvogSLEV, 1960):

Cr = xa~ (47)
where ~ is termed the "cohesion factor", and the failure condition may be written
as follows: (SKEMPTON and BISHOP, 1954):

(Ao.I __ Ao.3)f = xo-e cos 4br -I- (a e -- Auf)sin q$r (48)


1 - sin ~r
with ac = A a 3 f .
Since normally consolidated clays do not exhibit an effective cohesion inter-
cept, the corresponding failure condition reads (cf. eq.40):

sin 4b'
(Affl -- A63)f -- - - (tic - Auf) (49)
1 - sin 46'

It is an experimental fact that for clays, normally consolidated under various


ambient pressures and then sheared without drainage, the so-called "consolidated-
undrained" failure condition in terms o f total stresses can be expressed as:

sin ~b<,
l(Ao'l - Ao'3)f - o"e (50)
1 - sin ~b<<,

z'-/:/-~:~i, <..~ ~-~,/,~,-~/~/,


,4 T
,"~lc,,,,:::~:=,~,,'Tti:'-oilur'~ lll~Trtwe ,x-~,l~,'~

#
aea~,
i

Fig.18. Types of failure.

Earth-Sci. Rev., 6 (1970) 5~19


THE GEOTECHNICALPROPERTIES OF SOILS 31

The three types of failure discussed above are shown in Fig.18 from which it
follows, for example, that (cf. ALPAN, 1966):

Afn = N4" - N4'r - 2xx/N't'r (51)


2 z x / N 4 , r ( N O, -- 1)
and also:

Afn = co sin

where A f, = Auf/(A~ 1 -Ao-3) f is the pore pressure coefficient at failure for normally
consolidated clays (cf. eq.14).
A rather more generalized approach to the problem of shear strength has been
developed based on the concept of the so-called "critical state" (ROSCOE, 1967;
SCHOFIELD and WROTH, 1968).
I f the "state" of a sample is taken to be defined by the three variables a',
z and e (normal effective stress, shear stress and void ratio), the existence of a
"critical state line" in the (o-'-z-e)-space is postulated, representing the locus of
states in which the sample may undergo continuing distortion without concomitant

6-1 T

Fig.19. The critical state line. (After RoscoE, 1967.)

changes of stress or volume. In Fig.19 the position of the critical state line in the
(tr'-r-e)-space is shown as well as the two planes to which the drained and undrained
shearing tests are respectively confined.

Earth-Sei. Rev., 6 (1970) 5M-9


32 I. ALPAN

Time effects
It is an experimentally established fact that the shearing resistance of soils
increases at higher rates of deformation (CASAGRANDEand WILSON, 1951; WHIar-
MAN, 1957). This effect is particularly pronounced in clays and may be attributed
phenomenologically to a viscous component of their rheological response char-
acteristics. Thus, tests reported by CASAGRANDEand WILSON (1951) on a clay-shale
showed an almost two-fold increase in strength as the time to failure was reduced
from l min to 0.001 min.
Based on a relation proposed by TAYLOR (1948, section 15.12), the variation
of shearing resistance with strain rate, 8, may be written as:

s=-s0(1 + f(~) ) (53)


tan ~bo
where the subscript " o " indicates those values at which the strain rate is low
enough (say, of the order of 10-5/min) not to influence the shearing resistance, i.e.,
at which the function f(~) = 0. It is evident that f(~) increases monotonically
with 8. As a specific example, the empirical data reported by OHSAKI (1964) yield
the following expression:

f(~) = ~[2 - log(ef/~)] (54)


where c~ is a constant for a given clay and ef the strain at failure. An illustration
of additional factors influencing the relationship between strength and rate of
strain is afforded by a recent investigation by NAGARAJ (1968) on the role of soil
structure.
An aspect of an altogether different nature is involved when considering
the rate of shearing under drained conditions. The assumptions being here that at
all times during shear, only effective stresses are operative, it is necessary to ensure
a low enough rate for no pore water pressures to be set up.
The analysis of the problem, based on consolidation theory, yields the time
to failure, tf, at a given average degree of pore pressure dissipation, /.~f, (GIBSON
and HENI(EL, 1954); for example:
(1) Triaxial test; all-round drainage (cylinder 2R by 2H):

- (55)
tf 3/0~ -'[- 8fl 2 Cr(1---l-~f)
where: ~ = cr/c z and fl = H / R
(2) Direct shear (sample height = 2H):
H2
tf -- (56)
2Cz(1- Uc)
where Uc is the degree of pore pressure dissipation in the shear plane. Clearly, as
could have been anticipated, complete dissipation requires infinitely slow shearing.

Earth-Sci. Rev., 6 (1970) 549


THE GEOTECHNICAL PROPERTIES OF SOILS 33

S i n c e tf -~ gf/~ it is obvious that in order to stipulate the strain rate of a


particular test, information on the strain at failure is required. Experimental
evidence seems to indicate that f decreases with increasing overconsolidation and
might obey a relationship of the form (cf. SIMONS, 1960; BISHOP and HENKEL,
1962, table 8):

ef = a R -b (57)

where a = 2 0 - 2 5 ~ and b may vary between 0.4 and 0.8. R >i 1 is the "over-
consolidation ratio", a quantitative parameter of stress history and defined as the
ratio of maximum effective stress ever experienced to that applied in the test.

SELECTED SPECIAL TOPICS

Under this heading we shall discuss briefly some aspects of our survey
which, although important, have been the object of rather specialized inquiries.

Swelling soils
The spontaneous intake of moisture by cohesive soils is, in general, associated
with volume increase or, if the latter is prevented by appropriate confining con-
ditions, with the development of pressure.
The swelling processs in clays may be viewed as due, essentially, to osmotic
forces (WARKENT1Nand SCHOFIELD,1962), the extremely fine capillaries functioning
as a semi-permeable membrane. A broad classification of the factors influencing the
swelling characteristics of clay-water systems would be as follows (ALPAN, 1965b):
Qualitative factors:
Type of clay mineral
Texture of clay (composition)
Structure of clay (particle arrangement)
Quantitative factors:
Electrolite content of soil water
Exchange capacity (cation exchange)
Colloid content
Density
Moisture content
Degree of saturation
A somewhat simple model, assuming a regular array of clay platelets, was
analysed by WARKErqTIN(1962) who expressed the swelling pressure in terms of the
classical Van 't Hoff equation into which the cation and salt concentrations in the
pore water were introduced. This equation may be written as follows:

Ps = a(wo + b) -2 + c (58)
Earth-Sci. Rev., 6 (1970) 549
34 |. ALPAN

in which Wo is the initial water content a, b and c are constants for a given clay.
Four extreme points of the experimental curve of Fig.20 (KASSIFF and ZEITLEN,

i i

Colr'z~oc'ediC/~'f,"

~-xt~r/me~tfc,/
/(Unot,~/ur~o " Cl~fJ - -
1

I
I

-- ~v'~ (%) -

Fig.20. Swelling pressure vs. initial moisture content. (After KASSIFFand ZEITLEN,1961.)

1961) were used to calculate the constants appearing in eq.58 and the trend of the
resulting curve appears to be in fair agreement with empirical evidence, including
tests on compacted clay (WtSEMAN and ZE1TLEN, 1960).
Experimental evidence shows the time curves of the swelling pressure to be
similar to those of consolidation (NALEZNY and LI, 1967; BAKER and KASStFF,
1968) and may, therefore, be expressed by a function analogous to eq.35:

Ps=Po 1- M-~exp - ~ t (59)


n=0

with M = ~(1 + 2n)/2, Cps the appropriate coefficient of swelling pressure and
L a length representative of the flow-path geometry. The time-rate of swelling
pressure becomes, therefore:

dps
d~
=2po LCPs
~ ~
~ e x p ( - M2~cw
- t
) (60)
n=O

In practice, the time rate of swelling pressure is indeed a monotonously decreasing

Earth-Sci. Rev., 6 (1970) 5 4 9


THE GEOTECHNICAL PROPERTIES OF SOILS 35

function of time passing through zero when, as is often the case, the swelling
pressure exhibits a peak value (ALPAN, 1957).
A synoptic picture of the influence of various factors on the swelling pressure

~0

.
I
t.O

II
t.

- ~ - tsa..'o /"*/,.~ \

01
0 i'o ~o .3o -o ~o ~o 7"0

arnit/o/ Poroa/'~ "17o (o~)


Fig.21. Factors affecting swelling pressure. (After KASSlFF et al., 1968.)

of representative Israel clays is afforded by the empirical curves in Fig.21 (KASSIFF


et al., 1968, p.110) which may be conveniently expressed as follows:

dps
- - -- [(dwo - dwl)f(no) + ~dno] (61)
Ps

where f(no) is a linear function of the initial porosity, no, and positive provided
n o > 6 0 ~ ; Wo is the initial moisture content; w~ the liquid limit of the clay; and
~x a positive constant. It is evident that the swelling pressure decreases with in-
creasing moisture content and porosity and increases with the liquid limit, which
may be viewed as an over-all index of the physico-chemical characteristics of a
clay.
Additional empirical evidence, supporting the relationship expressed by
eq.61, has been presented by DANILOV0964) as shown in Fig.22. If we assume that
clays of equal swelling potential are represented in this figure by lines parallel
to the zone boundary, these may be expressed as:

Earth-Sci. Rev., 6 (1970) 5-49


36 T. ALPAN

fO0
fo
80 m
' 1 /
cloys.
~o

/
I
4o

~ 30
"4

"4
/5
/
fO
0 2O 4O

Zn/~/c,l Poro..e/7'q - n o {'I.)

Fig.22. Identification of swelling clays. (After DANILOV~1964).

dWl = q wl dn0 (62)

and compared with the condition of no change in swelling pressure and initial
moisture content imposed on eq.61:

dwl = [~/f(no)]dno (63)


If volume expansion during water intake is permitted, the process may be viewed,
phenomenologically, as reversed consolidation (TERzACHI, 1943, p.271). Since
the flow gradient during swell is directed into the clay, the pore water pressure
increases during the process and would be governed, for example in the one-
dimensional case, by the analogous equation (cf. eq.33):

dZu 3u
Csz - (64)
~z 2 ~t

where csz = coefficient of swelling.


Experimental swelling time curves have, indeed, the same appearance as
consolidation curves (e.g., DUBOSE, 1952; WISEMAN and ZEITLEN, 1960). Since the
effective stresses in the clay decrease during the swelling process, its volume change
characteristics are obviously those of an overconsolidated clay. Expressing the
change in void ratio, in analogy with eq.20, as:

Ae = - C~ log(a'/a'o) (65)

Earth-Sci. Rev., 6 (1970) 549


'THE GEOTECHNICALPROPERTIESOF SOILS 37

where Cs denotes the swelling index, it is reasonable to assume the index to be


influenced by the degree of overconsolidation.
Experimental evidence permits, indeed, a correlation between the indices
for normally overconsolidated clays in terms of the overconsolidation ratio
(ALPAN, 1966):

CJCc ~ k + 0.1 log R (66)


where, for plastic clays, the constant k has a value of the order of 0.2.

Thixotropy and sensitivity


Practically all cohesive soils are known to exhibit, after remoulding, a
strength increase with time. A similar phenomenon, characteristic of colloidal
suspensions, is that of thixotropy defined as the isothermal, reversible gel-sol

( -F/z-s/ 5oli#/f/dc'~'/'aM

~/-,9uefachon
. . . . I/r'n~,
a. Perfect Thtxo(ropy

/ I / . >--

~)'77e
b. .Zmperfoc/ Th/xo/ro,o f
Fig.23. Thixotropic behaviour. (After Bmuc, 1962.)

transformation produced in the suspensions by a mechanical disturbance (VON


ENGELHARDT, 1943; MITCHELL, 1961). In fig.23 time curves for perfect and imper-
fect thixotropy are shown (BILLIG, 1962).
Since many clays loose a considerable portion of their strength upon re-
moulding, it appears logical to assume a close connexion between this property,
termed sensitivity, and thixotropic behaviour.
By definition (SKEMPTONand NORTHLY, 1952), the Thixotropic Regain is:

Rt - ct - cr (67)
r

Earth-Sci. Rev., 6 (1970) 5~,9


38 1. ALPAN

where Cr is the remoulded strength and c t the strength measured after a certain
time of storage.
Denoting by cu the undisturbed strength, the quantitative expression for sensitivity
is:

S t = C~uu (68)
Cr
and the Remoulding Loss, L , can be defined as:

Lr - c u - cr _ St _ 1 (69)
Cr

A thixotropic recovery function can now be defined, which incorporates the


thixotropic time effects as well as the loss in strength due to the changes in the
clay structure caused by remoulding:

fe(t)--L,- -- 1 (St- 1) (70)

The recovery function evidently ranges from zero to unity (for perfectly thixo-
tropic materials). In general, however, ct < c. and thus the limiting value
fR (o~) < 1 is a measure of thixotropic imperfection. It appears, therefore, that
thixotropy cannot be considered as the cause of, at least, high sensitivity as
evident from the trend of relevant recovery functions (SKEMPTONand NORTHEY,

f~
ft7
r I I r r iill ]
Okuramur~ Cl~y

P
O,g

0 8 _ _ J

i
i I

- P/iit,Ir
iI

Io /oL7

3fo,'-/~ 77;'77e- d o u~
Fig.24. Typical thixotropic recovery. (After YAMAGUCHI,1959.)

1952). For clays of low or medium sensitivity (say, for S t ~< 8) the trend of.fRO)
indicates complete strength recovery with time as shown in Fig.24 based on data
reported by YAMAGUCm (1959).

Earth-Sci. Rev., 6 (1970) 5-49


THE GEOTECHNICALPROPERTIESOF SOILS 39

Plasticity, on the other hand, seems intimately related to thixotropy, both


being similarly influenced by characteristics of the clay-liquid system such as
specific particle surface, mineral composition, type and concentration of the
solution ions and the type of the liquid component. But whereas, with respect to
the geometry of the clay particles, thixotropy is a linear function of only the
specific surface, the level of plasticity is also influenced by particle shape (VON
PLATEN and WINKLER 1958).
Concerning the influence of thixotropy on strength and deformation char-
acteristics the following data taken at random from a considerable amount of
experimental work, must suffice as illustrations of typical trends:
NALEZNY and LI (1967) report a reduction of the amount of swelling and of
swelling pressure with storing time of compacted clays. They attribute this behaviour
to the re-formation of interparticle bonds during the thixotropic hardening with
the concomitant strength increase opposing swell (cf. MITCHELL, 1961). It may be
recalled that a similar relationship connects the Ko-coefficient with shearing
resistance.
SEED and CHAN (1957), working on compacted clays, examined the problem
of the influence of thixotropic hardening at different strain levels and moisture
contents. Fig.25 shows the variation of the stress ratio Pt/Po with axial strain at

~D

X8
/.7

/.6

N~ /.4

f~

Z/

,~xial S/rct/n - ~o

Fig.25. Thixotropic hardening of a compacted clay. (After SEEDand CHAN, 1957.)

high and low moisture contents, where Po is the stress applied immediately after
compaction and Pt the stress applied after storing--both stresses producing an
equal axial strain.
The physical basis of sensitivity does not, as yet, appear to be well under-
stood; all the same, a few speculative observations are in order. Remoulding imparts
to clay particles a certain degree of preferential orientation and the larger the

Earth-Sei. Rev., 6 (1970) 549


40 i. ALPAN

difference in orientation between the undisturbed and remoulded state of a clay,


the higher its sensitivity (MITCHELL, 1956). YAMAGUCHI (1959), applying the
theory of rate processes, arrived at an expression showing the sensitivity of a
clay to increase exponentially with the difference between the undisturbed and
remoulded activation energies.
The practical usefulness of the theory of rate processes in the study of the
mechanical properties of clays appears to me as not yet established, although
several relevant contributions have been made in this direction (cf. MITCHELL,
1964; MURAVAMA and SHIBATA, 1966; MITCHELL et al., 1968). Similarly, the
fracture of rocks has been analysed using the concept of activation energy heur-
istically (KtJMAR, 1968). It could be argued that, since changes in particle orien-
tation can be reasonably connected with changes in entropy, the observed depen-
dence of sensitivity on orientation might lead, in conjunction with the activation
energy relations postulated by YAMAGUCHI(1959), to a more consistent thermo-
dynamic formulation of sensitivity. Furthermore, as their parallel arrangement
corresponds to a stabler condition of the clay particles (MITCHELL, 1956), a
spontaneous return to an originally random structure with a corresponding strength
increase seems unlikely. Considerations of the kind outlined above would appear,
then, promising in explaining the relation between sensitivity and thixotropy.

Dynamic soil properties


The following discussion is concerned with the response of soils to dynamic
stimuli, i.e., the way their deformation characteristics are affected by the application
of rapidly changing forces. These may be of short duration (shocks), irregularly
fluctuating (earthquake- and blast-induced tremors) or periodically changing
(vibrations).
The kinetic energy, imparted to the soil during rapid loading, is partly lost
(irreversible deformations, heat) and partly radiated into the surrounding medium-
(for example as waves). From the viewpoint of energy transfer it appears con-
venient to classify the dynamic response of soils according to the level of their
energy states as follows (SLAOE, 1954):
(1) A high energy state in which changes in the average soil characteristics
occur; these are, essentially, changes in porosity.
(2) A medium energy state, characterized by irreversible local changes in
the soil structure; however, without significant changes in porosity.
(3) A low energy state in which the structural changes are reversible, i.e.,
elastic but not necessarily linear.
Each of these energy states is associated with a characteristic type of soil
response. Thus, the high energy state may result in compaction, rarefaction or
flow. In the medium state, energy is lost by dissipation or a process resembling
diffusion. In the low energy state a soil mass may be treated as an elastic conti-
nuum, provided the analysis proceeds from an elementary volume of suitable

Earth-Sci. Rev., 6 (1970) 5-49


THE GEOTECHNICALPROPERTIESOF SOILS 41

dimensions as compared with a given wavelength on the one hand and particle size
on the other.
The dynamic response of soils depends, in principle, on the relevant char-
acteristics of its constituents (solids and pore fluid), their relative mobility or degree
of "coupling" (cf. PATERSON, 1956) and on bulk parameters such as porosity,
degree of saturation, the structure of the particle skeleton, etc. In addition, the
response appears to be influenced by intergranular pressure and the type and
duration of the applied loading. It is, therefore, not surprising that the com-
plexities of the phenomena involved have, so far, precluded the formulation of a
reasonably integrated theory for real soils. All the same, the study of simplified
models on the one hand and extensive empirical investigations on the other, have
led to valuable insights and many important practical conclusions.
Thus, for example, WINTERKORN (1954) applied certain concepts from the
physics of the liquid state to granular assemblies at their critical void ratio, pre-
dicting their behaviour under the energy input associated with vibrations. L'HER-
MITE (1949) considered the grains to behave as simple resonators and viewed the
response of a grain assembly in terms of a velocity spectrum with respect to the
mass of the individual grains. Analysing a saturated model aggregate of spheres,
BRANDT (1955) derived an expression showing sound velocity to increase with
increasing effective stress and decreasing porosity, a trend in accordance with the
experimental finding of HARDIN and RICHART (1963).
Rheological models are often used to render complex material properties
amenable to analysis and the so-called "Kelvin-Voigt body" (or firmo-viscous
substance) has been found adequate in this respect (HARDINand SCOTT, 1966).
We shall use it here to illustrate the connexion between deformation and the rate
of load application.
The model consists, as is well known, of a "spring" and a "dash-pot"
coupled in parallel and its rheological equation (say, in axial stress) is:

~r = Ee + 24 (71)

where E = elastic modulus; 2 = Trouton's coefficient of viscous traction.


Consider a stress, ao, applied instantaneously and kept constant. We can
then define the "static stiffness", S~, as the ratio between the stress and the time
dependent strain, which yields:

S~ - ao _ E (72)
1 - e x p ( - t/Tret)

with Tret = 2/E = the retardation time.


Let, in a dynamic test, the applied stress and the resulting strain be periodic
functions of the form:

Earth-Sci. Rev., 6 (1970) 5-49


42 n. ALPAN

a = a0 sin cot (73)


e = e0 sin(ogt -- p)

and let, again, the "dynamic stiffness" be defined in terms of the stress and strain
amplitudes as:

Sd = a0/e0 (74)
Substituting these expressions in the rheological equation yields:

Sd = E[L + (2nTret/T)2] ~ (75)


where T = 2rc/~o = vibration period.
We can now define the "stiffness ratio", , as:

~P = Sd/Ss = [1 -- exp(-- t/Tre,)][1 + (2nTr~t/T)2] ~ (76)


and this comparative parameter evidently increases the longer the static testing
time, t, and the shorter the vibration period. Furthermore, as far as the intrinsic
material properties, as expressed by Tr~t, are concerned, the increase will be the

jLU
Io io B Io J /0 4

3 f ~tt~ I " l o d ~ le.,,~ off" .Z-I~,s//c/~q,

Fig.26. Dynamic and static moduli of elasticity.

more pronounced the more compressible the material and the stronger its viscous
component. The empirical curves of Fig.26 support the foregoing analysis.
Turning now to the closely related aspect of strength, it should be noted that
there exist two opposed tendencies as far as the effects of dynamic loading are
concerned. On the one hand, as pointed out elsewhere in this paper, the shearing
resistance of soils increases with the time rate of strain. On the other, the application
of dynamic forces, notably vibrations, tend to reduce the resistance to shear of
soils. Fig.27, based on results reported by SZAFRAN (1964), show the strength

Earth-Sci. Rev., 6 (1970) 5-49


THE GEOTECHNICAL PROPERTIES OF SOILS 43

~0
r"-,,
'k

f.o f .5 ~o ~.5

/ormol PreSSu~ , /~ - kq/Itm z

Fig.27. The effect of vibration on the strength decrease of a clay. (After SZAFRAN, 1964.)

/00
f ~

/
Dr',/~a,.~
:~ to

t~


Io0 500 zOO0 30oo

Fig.28. The effect of vibratory acceleration on the strength decrease of a dry sand. (After
MOGAMI and KuBo, 1953.)

decrease after vibration, As, relative to the pre-vibration strength, So, as a function
of normal pressure. Similarly, Fig.28, prepared from data for a sand reported by
MO6AMI and KuBo (1953), shows the influence of acceleration, the strength
being measured during the vibratory motion.
Qualitatively, the latter results may be explained on the basis of the concept
of "expansion pressure", postulated by L'HERMITE(1949) for fresh concrete and

Earth-Sci. Rev., 6 (1970) 5 4 9


44 1. ALPAN

applied by BA~ANT (1967) to sands, according to which the shear resistance of


a vibrated granular medium is practically zero below a certain limiting value of
confining pressure equal to the above-mentioned expansion pressure. The ex-
pansion pressure depends, amongst other things, on the kinetic energy supplied
to the medium, a conclusion evidently supported by the tests reported by MOGAM!
and KUBO (1953).
The dynamic model of a grain assembly consisting of individual resonators
responding within a given band width to excitation (L'HERMITE and TOURNON,
1948; L'HER~aITE, 1949) implies a threshold value for the input, below which the
shearing strength should remain practically unaffected. Recent vibration experi-
ments on sands (ST6TZNER, 1965) show, indeed, such threshold to exist at given
frequencies, velocities and accelerations.
We shall conclude our discussion of dynamic soil properties with some
remarks regarding their damping characteristics. Damping in soils is determined,
in many cases, by observing the decay of wave amplitudes with the distance from
the point of excitation. Part of the decrease in amplitudes is, of course, due to the
increasing volume excited as the waves radiate outward from the source, but in
part the dissipative properties of the material are responsible.
In the case of surface waves, for example, the following relationship may be
derived from the variation of energy density with distance:

Ar - V r ~ e x p [ - 2 ( r - r0) 1 (77,
A0
where Ar and Ao are the wave amplitudes at distances r and ro from the excitation
source and # is the "absorption coefficient" of the medium.
The absorption coefficient is, for many materials, dependent upon the wave
frequency. This dependence is shown in Fig.29, prepared from reported amplitude

/.= 3.I3 ~1o-4 To.zs

/j :
/
,L 4

I 1:5 2.0 2.5 d.O d.5 ~ 0

Fig.29. Energydissipation in seismic waves. (After ISSHtKIet al, 1962.)

Earth-Sci. Rev., 6 (1970) 549


THE GEOTECHNICAL PROPERTIES OF SOILS 45

measurements of tremors produced by the eruption of a volcano (IsSHIKI et al.,


1962).
In the laboratory, the methods used for measuring damping are mostly
based on the model of a linear resonator. Thus, HALL and RICHARF (1963) meas-
ured the decay in amplitude of sand specimens coming to rest in free vibration.
The damping was expressed in terms of the "logarithmic decrement", 6, defined
as follows:

A(t)/A(t + T) = exp(6) (78)


the ratio being that of amplitudes at times differing by a period. The logarithmic
decrement appears from the reported results to be amplitude-dependent, and may
be expressed as:

6 = mA" (79)
A similar trend in clays was reported by KONDNER and KRIZEK (1965) and we are
led to the conclusion that the stress-strain characteristics of the tested soils may
have deviated from linearity.

CONCLUDING REMARKS

It has been said of some books that they are never finished but have to be
abandoned by their authors. I am afraid that this applies with equal justice to
papers such as the present. Quite a few topics I would consider significant and
interesting have been left out, and probably many more considered as such by
other workers in the field. All the same, the survey presented here should afford
some insight into the fundamental problems of the study of the mechanical proper-
ties of soils and give an account of the methods applied in their attack.
No apology is offered for the uneven emphasis placed on the various topics
nor for altogether omitting a discussion of the thermal conductivity of soils, for
example. On the other hand, the control of soil properties or field and laboratory
testing are definitely subjects whose inclusion would have enhanced the usefulness
of the paper. However, a tolerably adequate treatment would have been pro-
hibitive which brings us back to the opening sentence of these remarks.

REFERENCES

ABOSIal, H. and MONDEN, H., 1960. An experimental method to determine the horizontal
coefficient of consolidation in fine grained soils. Proc. Reg. Conf. Soil Mech. Found. Eng.,
Asia, 1st, New Delhi, 12 pp.
ABosm, H. and MONDEN, H., 1961. Three-dimensional consolidation of saturated clay. Proc.
Intern. Conf. Soil Mech. Found. Eng., 5th, Paris, pp.559-562.
ALPAN, I., 1957. An apparatus for measuring the swelling pressure in expansive soils. Proc.
hltern. Conf. Soil Mech. Found. Eng., 4th, London, pp.3-5.

Earth-Sci. Rev., 6 (1970) 5 4 9


46 I. ALPAN

ALPAN, I., 1965a. Effective stresses in partly saturated soils. In: D. ABIR, F. OLLENDORFF and
M. REINER (Editors), Topics in Applied Mechanics. Elsevier, Amsterdam, pp.235-244.
ALPAN, I., 1965b. The Nature and Measurement of Volume Changes due to Moisture Variation in
Soils. Thesis, Israel Inst. Techn., Haifa, 66 pp. (unpublished).
ALPAN, I., 1966. The use of empirical relationships in evaluating the pore pressure coefficient A.
Bull. Intern. Inst. Seismology Earthquake Eng., 3: 39-51.
ALPAN, I., 1967a. The empirical evaluation of the coefficient Ko and Kog. Soil Found., 7 (1):
31-40.
ALPAN, I., 1967b. Notes on Soil Engineering. Intern. Inst. Seismology Earthquake Eng., Tokyo,
147 pp.
BAKER, R. and KASSIEF, G., 1968. Mathematical analysis of swell pressure with time for partly
saturated clays. Can. Geotech. J., 5 (4): 217-224.
BA~ANT, Z., 1967. Dynamic stability of saturated sand in subsoil beneath dams. Proc. Intern.
Congr. Large Dams, 9th, lstanbul, 4: 149-160.
BEAR, J., ZASLAVSKY, D. and IRMAY, S., 1968. 'Physical Principles of Water Percolation and
Seepage'. UNESCO, Paris, 465 pp.
BERNATZIK, W., 1947. Baugrund und Physik. S.D.V., Zurich, 310 pp.
BILLIG, K., 1962. Thixotropic clay suspensions and their use in civil engineering, 2. Civil Eng.
Public Works Rev., 57 (666): 101-105.
BLOT, M. A., 1941. General theory of three-dimensional consolidation. J. Appl. Phys., 12 (2):
155-164.
BISHOP, A. W., 1959. The principle of effective stress. TeA. Ukeblad, 39: 859-863.
BIsHoP, A. W. and BLIGHT, G. E., 1963. Some aspects of effective stress in saturated and partly
saturated soils. G~otechnique, 13: 177-197.
BISHOP, A. W. and ELDIN, G., 1950. Undrained triaxial tests on saturated sands and their sig-
nificance in the general theory of shear strength. Gkotechnique, 2: 13-32.
BISHOP, A. W. and HENKEL, D. J., 1962. The Measurement of Soil Properties in the Triaxial Test.
2rid ed., Arnold, London, 228 pp.
BOLT, G. H. and MILLER, R. D., 1958. Calculation of total and component potentials of water in
soil. Trans. Am. Geophys. Union, 39 (5): 917-928.
BRANDT, H., 1955. A study of the speed of sound in porous granular media. J. Appl. Mech., 77:
479486.
BRINCH-HANSEN, J. and LUNDGREN, H., 1960. Hauptprobleme der Bodenmechanik. Springer,
Berlin, 282 pp.
BUCKINGHAM, E., 1907. Studies on the movement of soil moisture. U.S. Dept. Agr., Bur. Soils,
Bull., 38:61 pp.
CAQUOT, A. and KI~RISEL, J., 1966. Trait~ de Mkcanique des Sols. 4th ed., Gauthier Villars, Paris,
506 pp.
CARRILLO, N., 1942. Simple two- and three-dimensional cases in the theory of consolidation of
soils. J. Math. Phys., 21 (1): 1-5.
CASAGRANDE,A., 1948. Classification and identification of soils. Trans. Am. Soc. Civil Engrs.,
113: 901-930.
CASAGRANDE,A. and WILSON, S. D., 1951. Effect of rate of loading on the strength of clays and
shales at constant water content. G~otechnique, 2: 251-263.
CASAGRANDE,L., 1952. Electro-osmotic stabilization of soils. J. Boston Soc. Civil Eng., 39 (l)"
51-83.
COLEMAN, J. D. and MARSH, A. D., 1961. An investigation of the pressure-membrane method for
measuring the suction properties of soil. J. Soil Sci., 12 (2): 343-362.
COOLING,L. F., 1945. Development and scope of soil mechanics. In: The Principles and Application
of Soil Mechanics. Inst. Civil Engrs., London, pp.l-30.
DANILOV,A. I., 1964. Diagram for dividing soils into ordinary, slumping and swelling. Soil Mech.
Found. Engrs., 5:324-326 (Transl. Consultant Bureau).
DERESIEWICZ, H., 1958. Mechanics of granular matter. In: H. L. DRYDEN and T. VON KARM.~N
(Editors), Advances in Applied Mechanics. Academic Press, New York, N.Y.,
5: 233-306.
DORSEY, N. E., 1940. Properties of Ordinary Water-Substance. Reinhold, New York, N.Y., 673 pp.

Earth-Sci. Rev., 6 (1970) 5-49


THE GEOTECHNICAL PROPERTIES OF SOILS 47

DRUCKER, D. C. and PRAGER, W., 1952. Soil mechanics and plastic analysis or limit design.
Quart. Appl. Math., 10 (2): 157-165.
DUBOSE, L. A., 1952. Evaluating taylor marl clay for improved use in subgrades. Texas Eng.
Expt. Sta., Res. Rept., 3 5 : 1 7 pp.
FORSLIND, E., 1952. A theory of water. Svenska Forskningsinst. Cement Betong Vid KgL Tek.
Hogskol. Stockholm, Handl., 16:43 pp.
FREUDENTHAL, A. M.. 1950. The Inelastic Behaviour of Engineering Materials and Structures.
Wiley, New York, N.Y., 700 pp.
FRYDMAN, S., 1968. The Effect of Stress History on the Stress-Deformation Behaviour of Sand.
Thesis, Israel Inst. Tech., Haifa, 170 pp. (unpublished).
GIBSON, R. E., 1953. Experimental determination of the true cohesion and true angle of internal
friction in clays. Proc. Intern. Conf. Soil Mech. Found. Eng., 3rd, Zurich, l: 126-130.
Gmsoy, R. E. and HENKEL, D. J., 1954. Influence of duration of tests at constant rate of strain on
measured 'drained' strength. G~otechnique, 4: 6-15.
GIBSON, R. E. and LUMa, P., 1953. Numerical solution of some problems in the consolidation of
clay. Proc. Inst. Civil Engrs., 1 (5877): 182-198.
HABIB, P. et SOEIRO, F., 1957. Les mouvements de l'eau dans les sols sous l'influence de latem-
perature. Cahiers Rech. Th~oret. Exptl. Mater. Struct., Assoc. Fran~. Rech. Essais, 56 pp.
HALL., J. R. and RICHART JR., F. E., 1963. Dissipation of elastic wave energy in granular soils.
J. Soil Mech. Found. Div., Am. Soc. Civil Engrs., 89 (SM6): 27-56.
HARDIN, B. O. and RICHART JR., F. E., 1963. Elastic wave velocities in granular soils. J. Soil Mech.
Found. Div., Am. Soc. Civil Engrs., 89 (SMl): 33-65.
HARDIN, B. O. and SCOTT, G. D., 1966. Generalized Kelvin-Voigt used in soil dynamics study.
J. Eng. Mech. Div., Am. Soc. Civil Engrs., 92 (EM1): 143-156.
HENNES, R. G., 1952. The strength of gravel in direct shear. Am. Soc. Testing Mater., Spec.
Tech. Publ., 131: 51-62.
HVORSLEV, M. J., 1960. Physical components of the shear strength of saturated clays. Am. Soc.
Civil Engrs., Res. Conf., Colorado, pp.169-273.
IRMAY, S., 1968. Failure criteria of plastic solids in the space of stress invariants. Israel J. Tech.,
6 (4): 165-173.
ISSHIKI, N., KUNO, H., MINAKAMI, T. and IWASAKI, I., 1962. O-Sima Volcano. Intern. Syrup.
Volcanol. (Inst. Architects Engrs., Intern. Union Geodesy Geophys.), Tokyo, pp.70-84.
JAKOBSON, B., 1957. Some fundamental properties of sand. Proc. Intern. Conf. Soil Mech. Found.
Eng. 4th, London, 1: 167-171.
JENNINGS, J. E. and BURLAND, J. B., 1962. Limitations to the use of effective stresses in partly
saturated soils. G~otechnique, 12: 125-144.
KASSIFF,G. and ZEITLEN, J. G., 1961. Pressures exerted by clay soil on buried conduits. Proc.
Intern. Conf. Soil Mech. Found. Eng., 5th, Paris, 2: 439-444.
KASS~rV, G., LlVNEH, M. and WISEMAN, G., 1968. Introduction to Pavements on Expansive Clays.
Technion Res. and Develop. Foundation, Technion I.I.T., Haifa, 197 pp. (in Hebrew).
K~zoI, A., 1966. Contributions to the investigations of granular systems. In: J. KRAVTCHENKO
and P. M. SIR1EYS(Editors), Rheology and Soil Mechanics - - Intern. Union Theoret. Appl.
Mech. Symp., Grenoble 1964. Springer, Berlin, pp. 164-178.
KLAUSNER, Y., 1964. Pure deviatoric loading of soils. Israel J. Tech., 2 (3): 305-311.
KONDER, R. L. and KRIZEK, R. J., 1965. Dynamic response of cohesive soils for earthquake
considerations. Proc. Worm Conf. Earthquake Eng., 3rd, New Zealand, 1 : 96-106.
KUMAR, A., 1968. The effect of stress rate and temperature on the strength of basalt and granite.
Geophysics, 33 (3): 501-510.
LAMBE, W. T., 1951. Soil Testing for Engineers. Wiley, New York, N.Y., 165 pp.
L'HERMITE, R., 1949. La rh6ologie du b6ton frais et la vibration. Rev. Mater. Constr., C, 405:
179-187.
L'HERMITE, R. and Totmyoy, G., 1948. La vibration du b6ton frais. Ann. Inst. Tech. Batiment
Tray. PubL, nouv. S&., 11:76 pp.
Liu, TH. K., 1967. A Review of Engineering Soil Classification Systems. Highway Res. Rec. 156:
1-22.

Earth-Sci. Rev., 6 (1970) 5-49


48 I. ALPAN

Low, PH. F., 1961. Physical chemistry of clay-water interaction. In: Advances in Agriculture, 13.
Academic Press, New York, N.Y., pp.269-327.
LvKow, A. W., 1958. Transporterscheinungen in kapillarpordsen Kdrpern. Akademie-Verlag, Berlin,
275 pp.
McK1NLAY, D. G., 1961. A laboratory study of rates of consolidation in clays with particular
reference to conditions of radial porewater drainage. Proc. Intern. Conf. Soil Mech. Found.
Eng., 5th, Paris, l: 225-228.
MICHAELS, A. S., 1952. Altering soil water relationships by chemical means. ,roe. Mass. Inst.
Tech. Conf. Soil Stability, Cambridge, Mass., pp.59-67.
MIKASA, M., 1965. The consolidation of soft clay. Civil Eng. Japan, Japan. Soc. Civil Engrs.,
pp.21-26.
MITCHELL, J. K., 1956. The fabric of natural clays and its relation to engineering properties.
Proc. Highway Res. Board, 35: 693-713.
MITCHELL, J. K., 1961. Fundamental aspects of thixotropy in soils. Trans. Am. Soc. Civil Engrs.,
126 (I): 1586-1620.
MITCHELL, J. K., 1964. Shearing resistance of soils as a rate process. J. Soil Mech. Found. Div.,
Am. Soc. Civil Engrs., 90 (SM1): 29-61.
MITCHELL, J. K., CAMPANELLA,R. G. and SINGH, A., 1968. Soil creep as a rate process. J. Soil
Mech. Found. Div., Am. Soc. Civil Engrs., 94 (SM1): 231-253.
MOGAMI, T. and Kuao, K., 1953. The behaviour of soil during vibration. Proe. Intern. Conf. Soil
Mech. Found. Eng., 3rd, Ziirich, 1: 152-155.
MURAYAMA,S. and SHIBATA,T., 1966. Flow and stress relaxation of clays. In: J. KRAVTCHENKO
and P. M. SIRIEYS (Editors), Rheology and Soil Mechanics - - Intern. Union Theoret. Appl.
Mech. Syrup., Grenoble 1964. Springer, Berlin, pp.99-129.
NAGARAJ,T. S., 1968. Strain rate influence on shear strength characteristics of a saturated kaolini-
tic clay. J. Mater., 3 (l): 210-235.
NALEZNY, C. L. and LI, M. C., 1967. Effect of soil structure and thixotropic hardening on the
swelling behavior of compacted clay soils. Highway Res. Records,, 209: 1-20.
NORTON, F. H., 1952. Elements o f Ceramics. Addison-Wesley, Reading, Mass., 247 pp.
OHSAKI,Y., 1964. On compressive strength of undisturbed cohesive soils under transient loading.
Proc. Architecture Inst. Japan, 98: 24-26.
PATERSON, N. R., 1956. Seismic wave propagation in porous granular media. Geophysics, 21 (3):
691-714.
RENDULIC, L., 1937. Ein Grundgesetz der Tonmechanik und sein experimenteller Beweis. Bau-
ingenieur, 18 (31/32): 459-467.
ROSCOE, K. H., 1967. Behandlung bodenmechanischer Probleme auf der Grundlage neuerer
Forschungsergebnisse. Bergbauwissenschaften 14 (12); 464-472; 15(1): 8-14.
RoscoE, K. H., SCHOFIELD, A. N. and WROTH, C. P., 1958. On the yielding of soils. GOotechnique,
8: 22-53.
ROSENQUIST, I. TH., 1961. Mechanical properties of soil-water systems. Trans. Am. Soc. Civil
Engrs., 126 (1): 745-779.
ROSENTHAL, D., 1964. Introduction to Properties o f Materials. Van Nostrand, Princeton, N.J.,
359 pp.
ROAD RESEARCH LABORATORY, 1952. Soil Mechanics for Road Engineers. Dept. Sci. Ind. Res.,
H.M. Stat. Office, London, 541 pp.
RUSSELL, M. B., 1942. The utility of the energy concept of soil moisture. Proc. Soil Sci. Soc. Ant.,
7: 90-94.
SCHOF1ELD,A. and WROTH, P., 1968. Critical State Soil Mechanics. McGraw-Hill, London, 310 pp.
SCHOEIELD,R. K., 1935. The pF of the water in soil. Trans. Intern. Congr. Soil Sci., 3rd, Oxford, 2:
37-48.
SEED, H. B. and CHAN,C. K., 1957. Thixotropic characteristics of compacted clays. J. Soil Mech.
Found. Div., Am. Soc. Civil Engrs., 83: (SM4) (1427): 1-35.
SHAW, D. J., 1966. Introduction to Colloid and Surface Chemistry. Butterworths, London, 186 pp.
SIMONS, N. E., 1960. The effect of overconsolidation on the shear strength characteristics of an
undisturbed Oslo clay. Am. Soc. Civil Engrs., Res. Conf., Colorado, pp.747-763.
SKEMPTON, A. W., 1954. The pore-pressure coefficients A and B. GOotechnique, 4: 143-147.

Earth-Sci. Rev., 6 (1970) 5-49


THE GEOTECHNICAL PROPERTIES OF SOILS 49

SKEMPTON,m. W., 1960. Significance of Terzaghi's concept of effective stress. In: From Theory to
Practice in Soil Mechanics. Wiley, New York, N.Y., pp.42-53.
SKEMPTON, A. W., 1961. Effective stress in soils, concrete and rocks. In: Conference on Pore Pres-
sure and Suction in Soils. Butterworths, London, pp.4-16.
SKEMPTON, m. W. and BIsHoP, A. W., 1954. Soils. In: M. REINER (Editor). Building Materials
Their Elasticity and Inelasticity. North-Holland, Amsterdam, pp.417482.
SKEMPTON, A. W. and NORTHEY, R. D., 1952. The sensitivity of clays. Gkotechnique, 3: 30-53.
SLADE JR., J. J., 1954. A discontinuous model for the problems of soil dynamics. Am. Soc. Testing
Mater., Spec. Tech. Publ., 156: 69-76.
SOUTO StLVEIRA, E. B., 1967. Thoughts concerning the applicability of the theory of elasticity to
soils. Proc. Panam. Conf. Soil Mech. Found. Eng., 3rd, Caracas, 1 : 3-27.
SOWERS, G. F., 1964. Strength testing of soils. Am. Soc. Testing Mater., Spec. Tech. Publ., 361:
3-21.
ST6TZNER, U., 1965. Ingenieurgeophysikalische Untersuchungen an Lockerb6den im Labor und
Gel/inde zur Bestimmung der Scherfestigkeit bei dynamischer Beanspruchung. Frei-
berger Forschungsh., C 158:58 pp.
SZAFRAN, Z., 1964. The effects of vibration on shearing resistance of clays. Proc. Seminar Soil
Mech. Found. Eng., Lodz, pp.513-519.
TAN, T. K., 1957. Structure Mechanics of Clays. Acad. Sinica, Inst. Civil Eng. Architects, Soil
Mech. Lab., Harbin, pp.l-17.
TAYLOR, D. W., 1942. Research on consolidation of clays. Publ. Dept. Civil and Sanit. Eng.,
Mass. Inst. Technol. Ser. 82, 147 pp.
TAYLOR, D. W., 1948. Fundamentals of Soil Mechanics. Wiley, New York, N.Y., 700 pp.
TERZAGHI, K., 1923. Die Berechnung der Durchl~issigkeitsziffer des Tones aus dem Verlauf der
hydrodynamischen Spannungserscheinungen. Sitz.-Ber. Deut. Akad. Wiss. Wien, Math.-
Naturw. KI., lla, 132 (3/4): 125-138.
TERZAGHk K., 1924. Die Theorie der hydrodynamischen Spannungserscheinungen und ihr
erdbautechnisches Anwendungsgebiet. Proc. Intern. Congr. Appl. Mech., 1st, Delft,
pp.288-294.
TERZAGHI, K., 1943. Theoretical Soil Mechanics. Wiley, New York, N.Y., 510 pp.
TERZAGHI, K. and PECK, R. B., 1967. Soil Mechanics in Engineering Practice. 2nd ed., Wiley,
New York, N.Y., 729 pp.
VAN PLATEN, H. und WINKLER, H. G. F., 1958. Plastizit~it und Thixotropie von fraktionierten
Tonmineralen. Kolloid-Z., 158 (1): 3-22.
VON ENGELHARDT, W., 1943. Zur Theorie der Thixotropie. Kolloid-Z., 102 (3): 217-232.
WARKENTIN, B. P., 1962. Water retention and swelling pressure of clay soils. Can. J. Soil Sci.,
42: 189-196.
WARKENTIN, B. P. and SCHOFIELD,R. K., 1962. Swelling pressure of Na-Montmorillonite in
NaCI solutions. J. Soil Sci., 13 (1): 98-105.
WmTMAN, R. V., 1957. The behaviour of soils under transient loadings. Proe. Intern. Conf. Soil
Mech. Found. Eng., 4th, London, 1: 207-210.
WILSON, G. and SUTTON, J. L. E., 1948. A contribution to the study of the elastic properties of
sand. Proe. Intern. Conf. Soil Mech. Found. Eng., 2nd, Rotterdam, 1: 197-202.
WINTERKORN,H. F., 1954. Macromeritic liquids. Am. Soe. Testing Mater. Spec. Teeh. Publ., 156:
77-89.
WINTERKORN, H. F., 1966. Shear resistance and equation of state for noncohesive, granular
macromeritic systems. Princeton Soil Eng. Res. Ser., 7 : 2 0 pp.
WISEMAN,G. and ZEITLEN,J. G., 1960. Swelling studies on a laboratory compacted clay. Proc.
Reg. Conf. Soil Mech. Found. Eng., (Asia), Div. 1, 1st, New Delhi, 16 pp.
YAMAGUCHI, S., 1959. On the sensitivity of clay. Disaster Prevent Res. Inst., Kyoto Univ., Bull.,
2 8 : 3 0 pp.
ZASLAVSKY, D., 1964. Saturated and unsaturated flow equation in an unstable porous medium.
Soil Sci., 98 (5): 317-321.

(Received May 7, 1969)

Earth-Sei. Rev., 6 (1970) 5~19

You might also like