The theory of one-dimensionalconsolidation of saturated clays.

11. Finite nonlinear consolidation of thick homogeneous layers

ROBERT
E. GIBSON
King's College, Lo~ldon,U.K., arid Golricr, Hoek and Associates Lrd., h4aide11Aead,U.K.

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

ROBERT
L. SCHIFFMAN
Departmer~tof Civil E~lgitzeeriilg,U~ziversityof Colorado, Borrlder, CO 80309, U.S.A.
AND

KENNETH
W . CARGILL
Waterrvays Experimet~tStation, Vicksbrrrg, M S 39180, U.S.A.
Received March 5 , 1980
Accepted January 7, 1981

The one-dimensional consolidation of a thick clay layer, initially consolidated fully under its
own weight, is considered. Account is taken of the variation of the coefficients of permeability
and compressibility as consolidation proceeds. T o render the theory consistent finite strains are
permitted. Comparisons with conventional theory, in a practical example, show that nonlinear
finite strain theory predicts the progress of consolidation settlement to be substantially swifter
than indicated by conventional theory, although the dissipation of excess pore pressure may be
slower. The consequences of this indicate that conventional consolidation theory has the
potential to seriously underestimate the excess pore water pressure in a soft layer. As a result,
the estimated shear strength would, if an effective stress analysis were used, be overestimated; a
potentially unsafe design could emerge.
On considere la consolidation uni-dimensionnelle d'une couche Cpaisse d'argile, initialement
consolide'e sous son poids propre. On tient compte de la variation des coefficients de perme'abilitC et de compressibilitC en cours de consolidation. De facon B rendre la theorie consistante,
des deformations finies sont autorisees. Les comparaisons avec la theorie conventionelle dans
un exemple pratique montrent que la th6orie en deformation finie non linCaire prCdit une
progression du tassement de consolidation beaucoup plus rapide que ce que donne la thCorie
conventionelle bien que la dissipation des pressions interstitielles puisse Ctre plus lente. Les
consequences de ce rtsultat indiquent que la theorie de consolidation conventionelle peut
6ventuellement sous estimer de facon importante la surpression interstitielle dans une couche
molle. Par consequent la resistance au cisaillement pourrait &tresur estimee dans une analyse
en contraintes effectives, conduisant B un projet potentiellernent non se'curitaire.
[Traduit par la revue]

Can. Geotech. J., 18,280-293 (1981)

Introduction
In an earlier paper (Gibson et (11. 1967) the equations governing the progress of one-dimensional
consolidation of a saturated clay layer were derived
on the basis of assumptions rather more general than
those usually adopted. It was shown that if the soil
skeleton is homogeneous1 and free from time-dependent effects (such as the ability to creep at constant
effective stress), then the void ratio e is a privileged
variable in the sense that in terms of it the governing
equations acquire their simplest form. The thin layer,
'The properties of the clay, such as permeability and compressibility, will, of course, alter through the thickness of the
layer and vary with time, owing to the changing distributions
of the void ratio; but this is nonhomogeneity of a derivative
kind. If the soil type, rather than merely its state, varies through
the layer, then the layer will be termed nonhomogeneous.

where the stresses arising from the self-weight of the

solids and pore fluid are negligible compared with
those applied, was examined in some detail and it
was shown that in these circumstances the void ratio
obeyed a weakly nonlinear equation of the parabolic
type; a result anticipated by McNabb (1960) and
Mikasa (1965).
In the present paper the case of a thick layer will be
considered. Clearly, the self-weight of the soil skeleton and the pore fluid must now be taken into
account, but the remaining assumptions adopted in
the earlier paper cited above will be retained. It will
be shown that the void ratio again obeys a nonlinear
equation, but the theory can be rendered linear under
conditions only slightly more restrictive than those
imposed to linearize the thin layer equation of the
earlier paper. Finally, this linearized equation is

0008-3674/81/020280-14$01.OO/O
@ 198 1 National Research Council of Canada/Conseil national de recherches du Canada

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

GIBSON ET AL.

solved nunlerically and the results recorded for a

problem likely to arise in practice.
We remark that although the load increment ratio
may be small and the void ratio changes and strains
small as a consequence, an unrestricted strain formulation of the problem is, nevertheless, required if the
stratum is to be treated as homogeneous. An alternative approach, valid for infinitesimal strains, is to
regard the medium as intrinsically nonhomogeneous,
even if the same clay type is present throughout
(Schiffman and Gibson 1964), but we shall not adopt
this viewpoint here.

Coordinate Systems
The usual coordinate system used in geotechnical
engineering is the ~ u l e r k nsystem where material
deformation is related to planes fixed in space. Thus,
the excess pore pressure in a consolidating clay layer
is measured at a point which is specifically related to
a fixed physical datum. It should be noted that under
this system the material particles (porous skeleton)
move with respect to the Eulerian coordinate system
as consolidation proceeds. Infinitesimal strain theories of consolidation assume that the thickness of
the compressible layer is constant; the deformation
of the layer, during consolidation, is assumed to be
small compared with its thickness. Using an Eulerian
system a piezometer located at some point in the clay
layer is referenced in space to a fixed datum. The
distance from the datum to the piezometer is always
supposed to remain the same.
Convective and Lagraizgian Coorcliizate Systems
If the deformations are large compared with the
thickness of the compressible layer, the use of an
Eulerian system can be very inconvenient. For example, a piezometer fixed in space but within the
zone of settlement, i.e. near the boundary of the clay
layer, may, after some time, be outside the layer! A
real measuring system is one which convects with the
material particles. The piezometer would always be
surrounded by the same material points and would
measure the pore water pressure of this part of the
skeleton as a function of its momentary position and
time. This type of system is termed a convective
coordinate system.
Consider the one-dimensional consolidation situation shown in Fig. 1 (Gibson et al. 1967). A clay layer
has an initial configuration as shown in Fig. la. This
is the configuration that would exist before consolidation begins. The bottom boundary of the layer can
be assumed to be fixed in space, e.g., a rock boundary
underlaying a soft clay. A thin sample of the clay
layer (AoBoC0Do)has a coordinate position a and has

a = a.

GRAVITY

8,

A,

.
...
...., -..,...

:I

<lo

D,
a

(a)

\'DATUM

PLANE

(:I

-01

(b)

DATUM PLANE

FIG.1. Lagrangian and convective coordinates: (a) initial

configu'atio" at r = 0 ; (b) configuration at time

thickness 60. Note that the position of the bottom

boundary (datum plane) is at a = 0. The position of
the upper boundary is at a = ao. The distance a is the
Lagrangian coordinate.
With consolidation the clay layer in the configuration shown by Fig. l a will have a new configuration
shown in Fig. lb. The datum plane remains fixed.
The top surface has moved and the sample has
deformed to a new position (ABCD). A new distance
C; locates a material point as a function of time; the
distance C; is the convective coordinate.
From a physical viewpoint it would be convenient
to express the dependent variables to be calculated
(such as excess pore pressure) in terms of the convective coordinate $, and time. This is mathematically
inconvenient since 4 itself is a function of the coordinate a and time t. The initial coordinate system a is,
however, independent of time and along with t is an
independent variable. Thus the mathematics of finite
strain consolidation is simplified by working in the
(a, t) system which is a Langrangian coordinate
system.
At first glance it might appear that the Lagrangian
and Eulerian systems are the same. It should be
noted, however, that whereas an Eulerian system is
fixed in space for all time, the Langrangian system
refers all events to an initial t = 0 configuration.
Events that occur with time may be related to either
the convective system (4, t) or to the Langrangian
system (a, 1). An example of the difference in use of
the convective and Langrangian system is visible in
plotting excess pore pressure isochrones. Using a
Lagrangian system in accordance with the notation
in Fig. 1, the isochrones would be plotted against a
fixed (in time) variable ranging from a = 0 to a = ao.
Using a convective system, the isochrones would be
plotted against a time-dependent variable ( ranging
from ( = 0 to C; = (0, where is a function of time.

282

CAN. GEOTECH. J. VOL. 18, 1981

A function F has a derivative with respect to a

defined by the chain rule as

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

But T and a are not related, thus

FIG.2. Change in void ratio during consolidation: (a) initial

state; (b) consolidating state.

Redztced Coordinates
Another set of coordinates, especially useful in the
present context, is based upon the volume of soil
particles lying between the datum plane and the point
being analyzed (Ortenblad 1930; McNabb 1960).
This coordinate z is defined as

[l]

z(a)

La

[ l - n(a', O)] da'

where n is the porosity. AS with the a coordinate, the

z labeling of a point is independent of time.
It turns out to be more convenient to work in terms
of the void ratio e, where
[2]

The derivative of 6 with respect to a has a physical

interpretation in terms of the void ratio e shown in
Fig. 2.
In the initial (reference) state at t = 0 an element
6a consists of solids (dimensioned as unity) and an
initial void ratio eo.At time t, the volume of solids is
still unity but the void ratio is e which is different
from eo. Thus the ratio of 6 t to 6a is

[g]

aF- l + e aF
aa
1 eo at

The derivative of a function F with respect to t is

e=n/(l-n)

and the reduced coordinate relationship of [l] becomes

da'
[31
z(a) =
o 1
e(al, 0)

where e is a function of a and t. ~ h u s

[lo]

a~

--

at

a~ at
a~ aT +-a~ at at at

From [5b] it is noted that the derivative of r with

respect to t is unity and so

dz/da = 1/(1 eo)

[4]
where eo is the void ratio at time t

0.

Coorclinate Transforrnrrtions
For reasons of physical clarity the derivation of the
governing consolidation relationships has been accomplished in the main in terms of the Lagrange
coordinate a, and here t(a, t) enters both as a dependent and an independent variable (Gibson et al.
1967). To insure that .$(a, t) enters only as a dependent variable, transformations between these two
systems must be established.
The relationship between a and $. is expressed
through the concept of the material derivative. Consider the convective coordinates (t, T) and the
Lagrangian coordinates (a, t). In the system described
above,
1501

The Governing Equation

We reintroduce briefly some results from the earlier
paper which will be needed in the development which
follows.
If the unit weight of the pore fluid pf and the solids
p, comprising the soil skeleton are both constant, it
has been shown that vertical equilibrium requires
that2

t = t(a, t)

and
[5b]

Equation [ll] expresses the material derivative. It is

noted that even though t and T are functionally equal
they must each be used in the appropriate context
and cannot be indiscriminately interchanged.

2The upper/lower sign in the ensuing analysis is taken if z is

measured againstlwith gravity.

283

GIBSON ET AL.

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

where o is the vertical total stress.

In addition, equilibrium of the pore fluid requires
that

wherep is the pore water pressure, and u is the excess

pore water pressure. If

The Finite Strain CoefJicient of' Consolidation

The function

plays the role of a coefficient of consolidation3 and it

is likely to be much less sensitive to changes in void
ratio than are its constituent terms. This suggests that
taking g to be constant may be a reasonable approximation for many clays, in which case [19] becomes

then continuity of pore fluid flow is ensured, where

v f and v s are the velocities of the fluid and solid

phases relative to the datum plane.

Finally, Darcy's law requires that

where p is the pore water pressure and k is the CQefficient of permeability.

If the soil skeleton is homogeneous and possesses
no creep effects and the consolidatian i s monotonic,
then the permeability k may be expected to depend
on the void ratio alone, so that
while the vertical effective stress

controls the void ratio, so thqt

Equations [12], [14], and [15] can now be united to
yield the following equation for the void ratio:

which appears to be highly nonlinear. This equation

can be rendered linear, while retaining the nonlinearity of the permeability and the compressibility,
by examining the relationship between soil properties.
It is noted that this equation is a general form of
the governing relationship for one-dimensional consolidation. It has been shown that both the conventional, linear, infinitesimal strain equation (Terzaghi
1924) and a nonlinear variation, which also assumes
infinitesimal strains (Davis and Raymond 1965;
Raymond 1969) are special cases of the above
(Schiffman 1980).

Using this approximation, the relationship between

the void ratio and the effective stress can be examined.
Relation between Void Ratio and EfSective Stress
Despite the simplification achieved by assuming g
to be constant, [19a] remains nonlinear by virtue of
the presence of the variable coefficient,

It is natural, now, to enquire into the consequences of

assuming h to be constant for then [19a] becomes a
linear equation.4 If this is the case, then the relation
between void ratio and effective stress implied by
[21] becomes
where coo and e, are the void ratios of the beginning
and end of consolidation, respectively. The form of
[22] is certainly compatible with the general shape of
the void ratio - effective stress relation encountered
with soils and has, indeed, been suggested previously
by a number of authors (see for example Zareteskii
1972) but the extent to which it can be regarded as an
adequate representation for real soils under load
must await further detailed examination.
Equation [22] is presented as a substitute for the
relation usually advocated to describe oedometer test
results for normally consolidated soils, namely
eo - C, loglo ( o ' / oo')
where eo is a reference void ratio, o 4 is the corresponding effective stress, and C, is the compression
index. Equations [22] and [23] are compared in Fig. 3.

[23] e

31f the usual coefficient of consolidation c, is used, however,

the term (1 e ) occurs in the numerator.
'Other approximations have considered the permeability
and compressibility of the soil to be functions of the excess pore
pressure (Koppula 1970) and the void ratio (Bromwell Engineering 1979). These approximations will not be considered
here.

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

CAN. GEOTECH. J. VOL. 18, 1981

"

od

Effective

Stress, o '

Fro. 3. Void ratio - effective stress relationships.

The 2 curves are fitted at a specific effective stress.

Thus a value of A is obtained for a measured value
of C,.
It is opined that [22]is a reasonable approximation
to [23]. It is worth remarking that 3 adjustable
parameters eoo,e,, and A are present in [22], while
only 2 are involved in [23].5Furthermore, it is clear
that a relation of the type [23] will for sufficiently
small or large values of u' sometimes raise difficulties,
although it is well established that it describes adequately the behavior of most clays over a limited
range of effective pressure.
One-dimensional Consolidation of Thick Layers
In the following section we shall examine the
progress of consolidation of a thick layer of homogeneous clay through which the parameters g and A are
constant. In these circumstances the void ratio obeys
the equation

a2e ae 1 ae
+
A(p, - pf) - =
[19b] az
az g a t

--

together with certain prescribed initial and boundary

conditions.
Consider a layer of initial thickness h(0) fully consolidated and in equilibrium under both its own
weight and a vertical effective stress q,' acting on its
upper surface as shown in Fig. 4. The layer is fully
saturated, with a water level at a distance H above
the top of the clay in this equilibrium state. The upper
surface of the clay layer is in contact with a pervious
soil. The lower surface of the clay can be in contact
with a pervious soil or may be impervious. We have
chosen to measure the reduced coordinate z down&Itis usual to write [23] with, apparently, 3 independent
parameters; the term (C,loglo ad) can, however, be absorbed
in eo, which reduces the number to 2

.. .<.I :.. -....,.;,. ..-. . ,* ' . Y

..'.,.,,,.....,

,::I-. r

'

PERV~OUS'" '

m
z=
X l

IMPERVIOUS

FIG.4. Thick, homogeneous clay layer.

ward from this upper surface. This surface serves as a

(moving) datum and so the positive sign must be
taken in [19b]above.
Layer with Perviow Upper and Lower Boundaries
We examine first the case where both the upper
surface z = 0 and the lower surface z = 1 are contiguous with pervious soils which are fully saturated
and into which the pore water from the clay can
escape freely. Since z is the volume of solids in a right
prism of unit (bulk) cross-sectional area between the
datum plane and a plane through the point P shown
in Fig. 4, the initial vertical effective stress there is
so that, by virtue of [22],the void ratio varies initially
through the layer according to the equation
[25]

e(z, 0) = ( ~ O-Oem)exp ( - Xqo' 4- (QS- pr)zI)

If the total vertical stress is augmented suddenly by

an amount Aq and maintained thereafter, this increased total stress immediately becomes fully effective on the boundaries z = 0, and z = I. In these
circumstances, therefore, the persistent boundary
conditions are

[27]

41, t )

( ~ O-Oem)
X exp ( - A[qd
=

+ (ps

pr)l+ Aqll

+ e,

Equations [25] through [27] in conjunction with

[19b]form a complete mathematical statement of the
problem.
Layer with Itnpervious Lower Bolindary
If pore water is free to escape from /he upper
r

GIBSON ET AL.

boundary, but the layer rests on an impervious base,

the analysis of the previous case may be carried over
in its entirety save that on the boundary z = I the
condition of impermeability requires that

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

[28]

ae
az

+ (PS- pi), dode

which boundary condition replaces ~271for this case.

Relation between Actual and Reduced Coordinates
All relationships governing the void ratio e above
are in terms of reduced coordinates. For practical use
to be made of these relationships, it is necessary to
relate the actual (Lagrange) coordinate a of a plane
of grains in the clay layer to the reduced coordinate z
of this plane. From [dl it follows that initially when
the a and z labellings are assigned,

where the initial vertical effective stress, ol(z, 0) is

given by 1241. By integrating [301 the following relation between z and a can be established,
[31]

s(t) [35]

U(t)

= --

S( )

vf = US

with this condition in [15], and subtracting the result

from [12], we find that when z = I,
[291

Furthermore

(1

+ e00)z + (~~( OP-OSe,)~ f )

X exp (-hq<){ 1 - exp [- hz(p, - pf)]1

where use has been made of the condition z = 0

when a = 0. At the boundary, z = I and a = h(0)
where h(0) is the initial thickness of the layer. A solution to this transcendental equation for 1 is readily
developed by simple computer-based techniques.
Thus both a and z are bounded and [31] relates
directly the a labelling to the z labelling.
The Settlement-Time Relation
The change in thickness 6S of an element of the
soil skeleton at time t is

L'

[e(z, 0) - e(z, t)] d l

J i[e(z, 01 - e(z,

m ;I]dz

which is the expression for the degree of settlement.

Effective Stress Isochrones
The solution of the appropriate boundary value
problem for e as a function of z and t can be accomplished by a variety of methods. Using these void
ratio results, the effective stress o' may be evaluated
from
1
4 2 , 2) - em
[22a] ul(z, t) = - - loge
h
Thus for every 6' at its coordinate position and time,
there is an
o'.
pore waterpressure
~~~~h~~~~~
The total stress is determined by integrating [12],
r;l

+ pr[S(t) + HI
where the sign is appropriate to the chosen positive
downward coordinate sense and account is taken of
the applied loads and the static pore water pressure.
The pore water pressure y is calculated by use of
the effective stress equation [17].
The pore fluid equilibrium relationship [13], namely

is used to calculate the excess pore water pressure u.

Non-dimensional Analysis
To simplify the numerical work we introduce at
this stage new variables
[38a] E(z, t) = e(z, t)/e(O, 0)
[38b] Z = z/l
[38c] T = gt/12

- pf)

so that the settlement S(t) of the whole stratum is

[384 N

Using [3] and [8]

in terms of which [19b] becomes

CAN. GEOTECH. J. VOL. 18, 1981

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

Normalized

Void Ratio,

E(Z,T)

Normalized

FIG. 5. Isochrones of normalized void ratio at different

degrees of consolidation.
Normalized Void Ratio,

E(Z,T)

Void

Ratio,

E(Z , T )

FIG. 7. Isochrones of normalized void ratio at 50% consolidation and different values of R.
Normalized

Void

Ratio,

E( Z,T)

FIG. 6. Isochrones of normalized void ratio at SOYo consolidation and different values of N.

FIG. 8. Isochrones of normalized void ratio at 50% consolidation and different values of B.

with the initial condition

If the boundary Z = 1 is impervious, then [42] is

replaced by
[43] a E / a Z + N ( E - B ) = O ;
Z = 1 , T>O
The numerical solution of [39],subject to conditions [40], [41.],[42], or [43], is a straightforward
exercise (Ames 1977; Cargill and Schiffman 1980).6

[40]

E ( Z , 0 ) = (1 - B ) exp ( - N Z )

+ B;
O I Z 5 1

and boundary conditions

[41]

E(O,T)=R;

[42]

E(l,T)=(R-B)exp(-N)+

if both boundaries Z

T>O

0 and Z

B; T > O

1 are pervious.

T h e computer program FSCONl which performs the

necessary calculations is available from the authors (Cargill
and Schiffman 1980).

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

GIBSON ET AL.

Time Factor, T

FIG.9. Non-dimensionalized settlement.

Figure 5 presents a typical example of the void ratio

isochrone for a layer which is impervious at the
bottom and pervious at the top. Figures 6-8 show the
effects of the variation of N , R, and B respectively.
A non-dimensional settlement S can be defined
such that
[44a] S(T) = le(0, O)S(t)
where

[44b] S ( T ) =

L1

[E(Z,0 ) - E(Z, T)] dZ

or, alternatively [44a] may be replaced by

The degree of consolidation U is then

Figure 9 presents the non-dimensional settlement

relationship for a range of values of N. Figure 10
presents the degree of consolidation relationship. It
is interesting to note that the degree of consolidation
is dependent only on the parameter N .
Equation [32] which expresses the relationship
between the Lagrangian a, and the reduced z coordi-

CAN. GEOTECH. J. VOL. 18. 1981

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

Dimensionless Reduced Coordinate,

FIG. 12. Dimensionless Lagrangian coordinates.

to h(0) and Z
[49],since
FIG.11. F ( 6 ) as a function of 6 for various values of Q.

1 , and hence a to be obtained from

[51] A = a/h(O) = F(NZ)/F(N)

Figure 12 shows the relationship between A and Z
as Q varies for a particular value of N.

TABLE
1. Soil properties-Osaka harbour mud

Plastic limit ( w , )
Liquid limit Ovl)
Water content at the top (rv)
Compression index (C,)
Coefficient of consolidation (c,)
Unit weight of water (pl)
Unit weight of solids (p,)
Total unit weight (p,)

nate can be rearranged as

[471

'(P' - pf)a = N Z
I
e,

+ Q[l - e x p ( - N Z ) ]

where the non-dimensional quantity Q is

More generally, at any point in the layer

N Z 4- Q [ l - exp (- N Z ) ]
a
A = -=
N
Q[l - exp (- N)l
h(0)
Figure 11 presents curves of the function

[49]

1501 ~ ( 3 - ) = 3- ~ [ -l~ X (-3-11

P
These enable N to be found from [26]setting a equal

Finite Strain Consolidation of a Soft,

Homogeneous Layer
We consider the application of the theory presented
herein to a typical field problem. This problem uses
the basic soil property data presented by Mikasa
(1965) for Osaka Harbour mud shown in Table 1.
A clay layer 10 m thick and pervious at both top
and bottom boundaries is considered in this example.
Initially, the water level is 10 m above the top surface
of the layer, and the initial vertical effective stress qo'
is zero. The effective stress at the lower pervious
boundary is initially 35 kPa.
Since the soil is saturated, the void ratio at the top
of the layer prior to consolidation is 3.83. This is the
chosen value of coo.
As part of a land reclamation project, 10 m of sand
is to be placed on the surface of the clay. The sand has
a dry unit weight of 1.61 t/m3 and a void ratio of
0.65. The sand will impose a loading Aq on the clay
of 98 kPa.
After consolidation is complete, the effective stress
at the lower boundary is 133 kPa, which, by [23],
corresponds to a void ratio e, of 1.49.
7A void ratio of 4.0 and an effective stress of 0.1 kPa are the
reference values in [23].

289

GIBSON ET AL.

0.4

2.

- 0

c 3

-8 E

8;-

.-

lF!

$2
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13
For personal use only.

0,3

-f S
-

02

0 U

0.1
0

5.0

Effectlve Stress,

100
a' (tsm)

15.0

FIG.14. Finite strain coefficient of consolidation as a function of the effective stress (1 tsm = 9.8 kPa).
Void Ratio, e
Effectlve Stress,

o' (tsm)

FIG.13. h - effective stress relationship (1 tsm

9.8 kPa).

The analysis of this practical example was developed by an explicit numerical procedure (Cargill and
Schiffman 1979).
The Coeficient (A)
The coefficient h is obtained from the relationship

where the void ratio e at the effective stress u' is

obtained from the logarithmic void ratio - effective
stress relationship. For the problem at hand the
exponential and logarithmic void ratio - effective
stress relationships will be matched at the average
effective stress a,'. Using values of coo and e, of 3.83
and 1.49, respectively, a plot of A versus u' is obtained, and is shown in Fig. 13. The average effective
stress is 67 kPa at a void ratio e, of 1.73 and we find
that h is 0.332.
Finite Strain Coeficient of Consolidation
The finite strain coefficient of consolidation g is
calculated by noting the relationship of g to c, previously mentioned. A reasonable approximation
would relate these 2 quantities as
which is shown in Fig. 14. The void ratio e, at the
average effective stress a,' will be chosen. Thus the
ensuing results will be based upon a value g of
0.00026 m2/day.
8The computer program FSCON2 which performs the
necessary calculations is available from the authors (Cargill
and Schiffman 1979).

FIG.15. Void ratio isochrones in terms of Lagrangian

coordinates (1 tsm = 9.8 kPa).

Void Ratio Isochrones

Using the above data a set of void ratio isochrones
are plotted at times representing 25, 50, and 75%
consolidation. The times are U = 0.25 at t = 264
days, U = 0.5 at t = 1164 days, and U = 0.75 at
t = 3018 days.
Figure 15 presents the void ratio isochrones in
terms of the initial a coordinates. Figure 16 presents
the same void ratio isochrones, but this is in terms of
the actual layer geometry at the times indicated. It
should be noted that in Fig. 16, at any particular
position in the clay, the material particles there will
alter in the course of time. This is not the case for
Fig. 15 where the same portion of soil skeleton is
identified for all time by a given value of the coordinate a. Therefore the small but real increase of
void ratio in the central portion of the layer is
different from the material present there prior to
consolidation; no real swell of the soil skeleton has
taken place.
It is further noted that initially the void ratio

290

CAN. GEOTECH. J. VOL. 18. 1981

E(Z,T)

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

Normollzed Void Ratlo,

1 .O

3.0

2.0

4.0

Void Ratio, e

FIG. 16. Void ratio isochrones in terms of the distance from

the bottom of the compressible layer (1 tsm = 9.8 kPa).

decreases with depth. This is a direct consequence of

the exponential void ratio - effective stress relationship which accounts for the self-weight of the deposit.
At the end of consolidation, the imposed stress Aq is
fully active and this value dominates, thus obscuring
the effect of the self-weight.
This example sets e(0, 0) = 3.83, and e(0, t ) =
1.5738. The dimensionless coefficients then become
N = 1.581, B = 0.389, R = 0.411, and Q = 0.940,
where I = 2.728 m. The values of the time factor at
25, 50, and 75% consolidation are T25 = 0.00922,
T60 = 0.04066, and T75= 0.10542. The dimensionless void ratio isochrones are plotted in Fig. 17.
Excess Pore Water Pressure Isochrones
The excess pore water pressure isochrones are
shown in Fig. 18. These isochrones appear to be
symmetrical about the momentary center of the layer.
They are, in fact, slightly skewed. In this example the
void ratios at the upper and lower boundaries are
close to each other, being 1.5738 and 1.5072 respectively. It is noted that the total stress due to the
submerged self-weight is (p, - pf)z so that at the
bottom of the layer (p, - pr)l = 46.55 kPa. This
compares with a total stress increment of 100 kPa
and we see that the effect of the application of the
total stress overwhelms the self-weight.
Settlement-Time Relationships
Figure 19 presents the settlement-time curves for
the top of the layer for a range of imposed loadings
ranging from a very light load of 10 kPa to a heavy
loading of 490 kPa. The magnitudes of the ultimate
consolidation settlement vary from 0.73 m for the
lightest load to 5.33 m for the heaviest load investigated. In each case the values of em, A, and g were
computed using the average effective stress and

FIG. 17. Normalized void ratio isochrones at 25, 50, and

75% consolidation.
10.0

aq.10 tsm

Excess Pore Water

t-0'

Pressure, u (tsrn)

FIG. 18. Excess pore water pressure isochrones in terms of

the distance from the bottom of the compressible layer
(1 tsm = 9.8 kPa).

corresponding void ratio. A value of 3.83 was chosen

for coo in all cases. These parameters are summarized
in Table 2.
Relation between Conventional and
Finite Strain Theory
A series of results have been presented which show
the predictions of finite strain consolidation theory.
It now remains to provide comparisons with conventional (Terzaghi 1924) theory. These comparisons
will be provided with respect to the degree of con-

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

GIBSON ET AL.

Time

lo3
t (days)

FIG.19. Settlement-time relationships for different imposed loading (1 tsm

9.8 kPa).

Time, t (days)

FIG.20. Degrees of consolidation for finite strain theory at various imposed loadings, and conventional theory (1 tsm

9.8 kPa).

TABLE
2. Finite strain consolidation parameters

solidation and the excess pore pressure isochrones, Degree of Consolidation

as these quantities are most familiar to geotechnical
Figure 20 presents the degree of consolidation U as
engineers.
a function of the time in days for imposed loadings at

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

292

CAN. GEOTECH. J. VOL. 18, 1981

Effective

Stress, a' (tsm)

FIG.22. Void ratio - effective stress relationships (1 tsm =

9.8 kPa).
Nornalized Excess Pore Water Pressure, u / ~ q

FIG. 21. Comparison of excess pore water pressure isochrones at 50% consolidation between conventional and finite
strain theories for moderate imposed loadings (1 tsm = 9.8
kPa).

1, 10, and 50 t/m2 (10, 100, and 500 kPa). Also

plotted on this figure is the degree of consolidation
which would be obtained by the use of conventional
theory. As expected, the finite strain theory predicts a
swifter degree of consolidation than would be indicated by conventional theory. For this case conventional theory gives 50% consolidation at 2521 days,
compared with 1164 days for a 10 t/m2 (100 kPa)
loading.
The effect of the magnitude of the loading is also
shown in Fig. 20. At early times in the consolidation
process the increasing load decreases the degree of
consolidation, with conventional theory forming an
upper bound. At later times the magnitude of the
imposed load takes precedence; the higher the load
the faster the consolidation process.
Figure 21 presents the normalized excess pore
water pressure, ~q'Aq,isochrones for both conventional and finite strain theory plotted in terms of the
distance from the bottom of the layer where the
settlement is zero. The imposed loading is 10 t/m2
(100 kPa). Although the degree of consolidation at
1164 days for conventional theory is 0.34 and is substantially less than that predicted by finite strain
theory, the excess pore water pressures are lower in
the upper and lower portions of the layer; the opposite is true in the central portions. This is contrary
to conventional wisdom which holds that standard
theory should provide higher excess pore water
pressures at lower degrees of consolidation.
The consequences of this indicate that conventional
consolidation theory has the potential to seriously
underestimate the excess pore water pressures in a

soft layer. As a result, the estimated shear strength

would, if an effective stress analysis were used, be
overestimated: a potentially unsafe design could
emerge.
The reasons for this difference in excess pore water
pressure lies in the void ratio - effective stress behavior. This is shown in Fig. 22 for 3 cases. First, the
conventional theory is represented by a straight line
governed by
where a, is the oedometer compressibility. Normal
practice defines a, by
The other curves in Fig. 22 show the exponential
and the logarithmic void ratio - effective stress relationships. It is noted that the value of a, is quite small
and the approximation to reality is poor. To provide
a more realistic approximation using infinitesimal
strain theory, the value of a, should be increased.
This in turn would decrease the coefficient of consolidation. Then the degree of consolidation at 1164
days would decrease. This means that the excess pore
water pressures would increase.
The above phenomenon holds for all but the very
lightest load where the self-weight of the compressible layer dominates the process of consolidation as shown in Fig. 23. Here it is noted that
conventional theory will underestimate the excess
pore water pressure at the bottom of the layer. The
reverse is true at the top of the layer.
Thus it is clear that the normal use of conventional
theory can overestimate the pore pressure dissipation
due to its use of linear, infinitesimal strain approximations. Furthermore, the softer the soil, and the
greater the loading, the greater the overestimation of
pore pressure dissipation.

Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Michigan State University on 05/21/13

For personal use only.

GIBSON ET AL.

293

Acknowledgements
We are pleased to acknowledge the cooperation
and advice of Dr. G. L. England of King's College,
London, and Professor M. J. L. Hussey of the Open
University, United Kingdom, at an early stage of
this work. We are further pleased to acknowledge the
partial financial support provided by the National
Science Foundation to the U.S. authors.

Nwmallzed

Excess Pore Water Pressure, u / ~ q

FIG. 23. Comparison of excess pore water pressure isochrones at 75% consolidation between conventional and finite
strain theories for very light imposed loading ( I tsm = 9.8
kPa).

Conclusions

In this paper we have presented a consistent theory

of finite strain consolidation which is applicable to
loaded, thick, homogeneous clay layers. The effect of
the self-weight has been examined. A comparison
with conventional consolidation theory has shown
that conventional theory will seriously overestimate
the time of consolidation but can underestimate the
amount of excess pore pressure at a given time. This
latter situation can lead to an overestimation of the
shear strength of a deposit.
The theory as cast above is particularly advantageous in engineering practice, since its use does not
require any special tests. Conventional oedometer
data can be used to provide adequate results
In spite of its advantages it must be noted that the
theory remains an approximation in several important respects: first, it is assumed that the finite strain
coefficient of consolidation is a constant throughout
the process of compression. This is clearly not the
case for real soils. Secondly, the soft deposits which
are likely to develop appreciable strains and for which
the use of a finite strain theory is mandatory are also
unlikely to be free from intrinsic time effects such as
skeleton creep. Furthermore, the presence of nonhomogeneity is clearly the rule rather than the
exception. These factors have been ignored explicitly
in the development, but not in the formulation of the
theory. A11 these assumptions in the context of other
situations such as marine sedimentation we leave to
a later paper.

AMES,W. F. 1977. Numerical methods for partial differential

equations. 2nd ed. Academic Press, New York, NY, 365 p.
BROMWELL
ENGINEERING.
1979. Analysis and prediction of
phosphate clay consolidation: Implementation package.
Bromwell Engineering, Lakeland, FL.
CARGILL,K. W., and SCHIFFMAN,
R. L. 1979. FSCON2-I,
version 1, level A, one-dimensional finite strain consolidation
of a thick normally consolidated homogeneous layer with a
constant coefficient of consolidation. Geotechnical Engineering Software Activity, Department of Civil Engineering,
University of Colorado, Boulder, CO.
CARGILL,
K. W., and SCHIFFMAN,
R. L. 1980. FSCON1-I,
version 1, level A, one-dimensional finite strain consolidation, dimensionless void ratio for a thick, normally consolidated homogeneous layer. Geotechnical Engineering
Software Activity, Department of Civil Engineering,
University of Colorado, Boulder, CO.
G. P. 1965. A non-linear theory
DAVIS,E. H., and RAYMOND,
of consolidation. Giotechnique, 15, pp. 161-173.
G. L., and HUSSEY,
M. J. L. 1967.
GIBSON,R. E., ENGLAND,
The theory of one-dimensional consolidation of saturated
clays, I. Finite non-linear consolidation of thin homogeneous
layers. Geotechnique, 17, pp. 261-273.
KOPPULA,
S. D. 1970. The consolidation of soil in two dimensions and with moving boundaries. Ph.D. dissertation,
Department of Civil Engineering, University of Alberta,
Edmonton, Alta.
MCNABB,A. 1960. A mathematical treatment of one-dimensional soil consolidation. Quarterly of Applied Mathematics, 17, pp. 337-347.
MIKASA,
M. 1965. The consolidation of soft thy, a new consolidation theory and its application. Japanese Society of
Civil Engineers (Reprint from Civil Engineering in Japan,
1965), pp. 21-26.
ORTENBLAD,
A. 1930. Mathematical theory of the process of
consolidation of mud deposits. Journal of Mathematics and
Physics, 9(2), pp. 73-149.
RAY~IOND,
G. P. 1969. Consolidation of deep deposits of
homogeneous clay. Geotechnique, 19, pp. 478-494.
SCHIFFMAN,
R. L. 1980. Finite and infinitesimal strain consolidation. ASCE Journal of the Geotechnical Engineering
Division, 106(GT2), Proceedings Paper 15193, pp. 203-207.
SCI- IFFM MAN, R. L., and GIBSON,R. E. 1964. Consolidation of
nonhomogeneous clay layers. ASCE Journal of the Soil
Mechanics and Foundations Division, 90(SM5), Proceedings Paper 4043, pp. 1-30.
K. 1924. Die Theorie der Hydrodynan~ischen
TERZAGHI,
Spannungserscheinungen und ihr Erdbautechnisches Anwendungsgebiet. Proceedings, First International Congress
of Applied Mechanics, 1, Delft, Netherlands, pp. 288-294.
ZAKETSKII,
Yu. K. 1972. Theory of soil consolidation. (Translated from Russian by the Israel Program for Scientific
Translations.) U.S. Department of Commerce, National
Technical Information Service, Springfield, VA, 308 p.