Consolidation
S. HANSBO,*
by vertical drains
M. JAMIOLKOWSKIt
A theory is described which can be used in designing a
vertical drainage system. This theory incorporates
important parameters
such as vertical discharge
capacity,
remoulding effects during installation and filter resistance.
The importance of these parameters is widely recognized,
but very seldom taken into account in the design, by
engineers. Six well-documented
case records, related to
different drain types are briefly presented.
Experience
gained from these case records shows that in many cases,
even if the overall performance
appears successful, there
are aspects of soil/drain
behaviour
that cannot
be
explained on the basis of existing theories.
and L. KOKS
plasticity index
permeability of undisturbed
soil
permeability of soil in zone of smear
axial permeability of drain
half length of drain
Did
discharge capacity of the drain well
pore pressure coefficient
&Id
thickness of band-shape drain
time factor in radial consolidation
excess porewater pressure
average degree of consolidation
natural moisture content
liquid limit
plastic limit
depth of soil
bulk density
change in porewater pressure
principle consolidation
stress
vertical preconsolidation
pressure
original effective stress
Larticle dtcrit une theorie que lon peut appliquer a la
conception
dun systeme de drainage vertical Cette
theorie renferme dimportants parametres tels que la
capacite de decharge verticale, les effets du remaniement
pendant
linstallation
et la resistance
de filtration.
Limportance de ces paramttres
est reconnue par bien des
ingtnieurs, mais ils en tiennent rarement compte au stade
de la conception. Larticle donne une description
sommaire de six cas bien document&s dans lesquels divers
types de drains sont utilises. Lexperience acquise grace a
cette etude de cas montre que, tres souvent, m&me si
lellicacite globale semble satisfaisante, il existe certains
aspects du comportement
sol/drains
que lon ne peut
expliquer sur la base des theories existantes.
THEORETICAL CONSIDERATIONS
The design of a vertical sand drain system is
generally based on the classical theoretical solution
developed by Barron (1948) in which the drains are
assumed to be functioning as ideal wells, i.e. their
permeability is considered infinitely high as compared with that of the soil in which the drains are
placed. This assumption is justified when the drain
sand fulfils the requirements
of an ideal filter
material.
However,
in practice
it is doubtful
whether such an ideal condition can be achieved. If
the permeability
of the sand is in the order of
magnitude of 500-lOOOm/year, the effect of well
resistance cannot be ignored.
In the case of prefabricated
drains the effect of
well resistance must be taken into account, particularly when the drains are long. A simple
theoretical
solution to this problem is needed.
Solutions which take into account the effect of well
resistance have existed for some time (Barron, 1948;
Yoshikuni & Nakanodo,
1974) but although these
solutions are based on the simplified assumption of
equal strain they are still very complicated to use in
practice. For example, Barrons solution in respect
NOTATION
cross-sectional
area of the drain well
effective cohesion
width of band-shape drain
compression
index
coefficient of consolidation
in the horizontal direction
unconfined compressive strength
coefficient of consolidation
in the vertical
direction
diameter of soil cylinder
diameter of drain
equivalent diameter
diameter of zone of smear
original void ratio
undrained modules of deformation at 50%
failure stress
*Chalmers Tekniska Hiigskola, Giiteborg
t Politecnico di Torino
j Openbare
Werken
Amsterdam.
45
S. HANSBO,
M. JAMIOLKOWSKI
AND L. KOK
Distance from centre of dewatered cylinder with
radius R = 0.5 m
Fig. 2 Comparison of consolidation effects caused by a
100 mm x 4 mm band-shaped drain and a circular drain
with equivalent circumference; d = 66 mm, D = 1 m, time
factor q = q, t/D2
s = d,/d
qw = k, A, k, nd2f4 is the discharge
the drain well.
Fig. 1. Schematic picture of soil cylinder dewatered by
vertical drain
0, = 1 -exp(
-ST,/&
(1)
where
Th = Cht/o * is the time factor
consolidation
pS = In (n/s)+(k,/k,)
In(s) - 3/4
+ rrz(21-z) kc/q,
n = D/d
in
radial
of
The derivation
of this formula is presented by
Hansbo, 1981). Equation (1) is deduced for a drain
well with circular cross-section.
A band-shaped
drain has to be converted into a circular cylindrical
drain producing the same consolidation
effect as
the band-shaped
drain. The equivalent diameter d,
of this drain (cf Fig. 2) is
=2(b+t)
of well resistance is given in implicit form and the
diagrams presented
are not correct (Fig. 10 of
Barron, 1948). Yoshikuni & Nakanodos solution
includes Bessel functions of the zeroth and first
order and thus requires the use of mathematical
tables or advanced computers. However, a simple
solution to the problem of smear and well resistance was presented
by Hansbo (1979), giving
results almost identical with those presented by
Barron and Yoshikuni & Nakanodo.
Thus for a water-saturated
soil the average
degree of consolidation
(C7,,)at a depth z due to the
effect of radial drainage
only (Fig. 1) can be
expressed as
capacity
where b is the width of and t is the thickness of the
band-shaped
drain (Hansbo, 1979).
This discharge
capacity of the prefabricated
band-shaped
drains will be a function of the
effective lateral earth pressure against the drain
sleeve. In the majority of the cases the filter will be
partly squeezed into the channel system of the core
by the pressure of the surrounding soil and this will
reduce the channel volume and consequently
the
discharge capacity. The discharge capacity may
also decrease with time. For example, fines in the
pore water may not be retained by the filter and
may cause a gradual decrease of discharge capacity
or even clogging. The filter permeability should not
be higher than required with respect to the discharge capacity. The filter permeability
of the
existing prefabricated
drains (which according to
the results of laboratory
tests have short-term
discharge capacities of maximum 10-25 m/year)
need not be higher than 0.01-0.05 m/year (Hansbo,
1981).
VERTICAL
Average degree of consolidation:
0
40
shaped drains are used; these have an equivalent
diameter of 0.062 m and discharge capacities of 5,
10 and 20 m3/year. In the other case displacementtype sand drains are used; these are 0.18 m in
diameter, with discharge capacities of 13 m3/year
(permeability
of
drain
sand = 500 m/year
= 1.6 x 10e4 m/s) and 130m3/year. In both cases
the zone of smear is assumed to have a diameter
d, = 2d and a permeability k, = 1/3k,. The clay in
which the drains are placed has a permeability
k, = @03 m/year in the horizontal direction. The
coefficients of consolidation
are c,, = 0.5 m*/year
and c, = @15 m/year. The result of the analysis in
respect of two different spacings is given in Table 1
and Fig. 3.
Whether the drains are open or closed at the
bottom has an important influence on well resistance. Well resistance of penetrating drains 30 m in
length would have the same negative effect as that
of drains 15 m in length which are not penetrating.
Moreover, the tip of the band-shaped
drains is
generally folded round an anchor rod during
installation. Therefore even penetrating drains may
have the bottom outlet in clay and thus be more or
less closed unless they are driven to a certain depth
in pervious soil.
A more detailed theoretical parametric study has
been made by Hansbo (1981). However, the best
and safest basis for a comparative
study is to use
data from a large number of full-scale tests in which
the effect of different drain types can be directly
compared. As yet the Authors have not enough data
to provide a statistical basis for such a comparison.
80
f
7-i
jl
Prefab ban&shaped
--.-
Sand drain, d=
drain
qw = 20 mslyear
qw = 10 ms/year
qw = 5 ms/year
47
DRAINS
0.18 m
kw = 5000 m/year
kw = 500 m/year
Fig. 3 Influence of well resistance on average consolidation at di5erent depths in clay with c, = O.lSm/year,
q, = 05 m/year
and k, = 0.03 m/year; drain spacing
@9 m (D = 095 mk discharge capacity qW= k, A,
Drains with low filter permeability ought to be
cut as close as possible to the groundwater
level in
pervious soil or in the drainage blanket. Otherwise
the water level inside the filter, in the drain core,
may cause excess back-pressure
in the wells, thereby reducing their efficiency.
For a certain fixed value of s, the influence of
smear will increase with increasing diameter of the
drain. However, the s value may be greatly influenced by the installation method. Equation (1)
offers the possibility of easily making a systematic
study of the influence of smear and well resistance.
Take as an example two typical cases with 30 m
long fully penetrating drains (draining at top and
bottom). In one case prefabricated
and band-
COMPARATIVE
CASE RECORDS
Despite
the previously
outlined
theoretical
developments vertical drain design is still subject to
many uncertainties;
these will be overcome when
more experience is gained in respect of field performance related to various soils all over the world.
The results of six geotechnically
well-documented
case records are presented here, with the aim of
evaluating how factors such as soil parameters,
drain types, spacing and installation
procedures
Table 1. Average consolidation (%) at 15 m depth without and with regard to smear and well resistance; drain
spacing = 1.5m (D = 15gm);
discharge capacity of prefabricated drain = 20m3/year
and of sand
drain = 13 m3/year
I
Time of
consolidation:
years
No smear
ideal
Smear
ideal
wells
wells
0.5
1
42
67
89
99
25
44
68
90
2
4
Prefabricated drains
Sand drains
Smear
well
resistance
17
31
52
77
No smear
ideal
Smear
ideal
wells
wells
27
47
72
92
19
34
56
81
Smear
well
resistance
15
28
48
73
t c, according
10
Depth:
m
?!
w
Brown grey
varved
Bands of
iron sulphide
Grey coloured
by iron
sulphide
Green-grey
slightly
organic
to field vane test.
3
3
M
z
a
Dry crust
Area V
Geodrains
s1
3
2
M
z
z
Soil profile
Grey varved
Bands
of iron
sulphide
Grey coloured
by iron
sulphide
Grey-green
Dry crust
Area I
Sand drains
1.61
1.68
1.62
160
1.66
1.66
1.59
1.43
1.62
1.60
1.62
1.53
1.46
1.28
1.45
I.58
1.48
1.48
1.55
y: t/m3
66
67
68
79
65
60
72
103
63
70
68
82
89
128
100
64
92
90
II
95
w: y0
WL:
79
48
50
50
58
16
78
61
16
21
65
20
22
22
27
25
22
23
28
25
62
wp: %
20
41
61
58
85
63
72
84
150
132
/,
60
65
75
70
64
63
28
CF
0.33
0.30
0.30
0.31
0.32
0.34
0.34
0.35
0.50
0.22
0.26
0.32
0.30
0.32
044
0.93
2.2
0.2 1
0.23
tUL3
1.1
1.1
1.2
1.1
2.0
OCR
1.2
1.2
1.5
ground water level at 05 to 1 m depth; hydrostatic condition; studied
106
Table 2. Geological profile and geotechnical characteristics of soil in test areas I and V at Sk&Edeby;
layer = 2.5 to 75m
50
S. HANSBO,
M. JAMIOLKOWSKI
AND
L KOK
o Bellows hose
30 m
Spacing I.4 m
sand drains
813 814
00
Spacing 1.6 m
sand drains
Spacing 1.1 m
sand drains
B15
812
I.-.-.
.-.
.~.
.-
I-
--
83 84
00
85 B6
00
87 88
00
&ieodraii
Alidrain
10
12
, IAlidrain
Undrained shear strength, cu: kPa
2-
6-
I
Sand drain (I)
0.9 m spacing
Fig. 7. Increase in undrained shear strength
consolidation; test areas I and V at Sk&Edeby
&O
Geodrainl
Bl?
0812
Alidrain
Soft clay
Moraine
Soft clay
Fig. 6 Excess porewater pressure (at 5m depth) and
settlement (compression of clay layer between 25 and 75 m
depth) vs time in the centre of test areas I, IV and V at Sk&
Edeby
8;
14
Time: years
Geodrain (V)
0 9 m spacing
,Bl8
el6
due to
of the Geodrain. Samples of Geodrain were taken
in test pits to depths below the groundwater
level
before the additional load was applied.
The results of the settlement and pore pressure
observations
are given in Fig. 6. The rate of
settlement in Area I with sand drains was greater
than in Area V with Geodrain. The efficiency of
Geodrain was no less with time although the filter
paper, collected from the test pit after 3f years, was
partly degraded. The back-calculated
coefficient of
consolidation,
giving the best fit with the observed
rate of settlement, has the same value, both after
Moraine
Section I
Fig. 8.
Plan and cross-section of teat field at brebro;
depth of clay lay& varying between about 8 and 9.5 m
and before the additional loading. Comparing the
rate of settlement for the two areas the effect of
smear seems more pronounced for the sand drains
than for the Geodrains. A good correspondence
between theory and practice would be obtained
assuming the values c,, = 0.5 m/year, s = 2 and
kc/k, = 3 for the sand drains and s = 2 and
k,/k, = 2 for the Geodrains. The increase in undrained shear strength caused by consolidation
in
the drained areas is presented in Fig. 7. In Area V
the increase is less pronounced
at greater depth
indicating the influence of well resistance.
The result of pore pressure measurements
in
Area I shows how difficult it may be to judge, from
pore pressure readings, the degree of consolidation
achieved. In Area V there still remains an excess
pore pressure of about 15 kPa in the centre of the
studied layer.
&ebro case record
In connection
with the construction
of a new
motorway section between Grebro and Gothenburg a test field, 125 m by 45 m, was arranged just
outside Grebro for the purpose of comparing the
efficiency of the two types of prefabricated
drains
manufactured
in SwedenGeodrain
and Alidrain-with
that of displacement-type
sand drains,
0.18 m in diameter. The test field was divided into
three sections of equal size with drain spacings
1.4 m, 1.1 m and 1.6 m in an equilateral triangular
pattern (Fig. 8).
VERTICAL
51
DRAINS
Undrained shear strength cu: kPa
Spacing 1.1 m
11
11
17
16
21
18
30
24
23
24
41
51
43
27
50
81
Spacing 1.6 m
Spacing 1.4 m
S, sensitivity
(1) area with sand drains
(2) areas with Geodrain and Alidrain
Fig. 9. Original undrained shear strength in the &ebro
test field determined by the Swedish fall-cone test
The geotechnical data of the soil are presented in
Fig. 9 and Table 3. As in the case of Sk&Edeby, the
soil characteristics
vary between the three studied
sections. The consolidation
characteristics
according to oedometer tests are also subjected to variations which make the interpretation
difficult.
The average
virgin
compression
modulus
increases with depth from an average of about
CR = 0.4 just below the dry crust to an average of
about CR = D65 in the lower part of the clay layer.
In a few cases values of up to CR z 1 have been
measured, indicating some kind of structural collapse when the preconsolidation
pressure
is
exceeded.
The coefficient of consolidation
in the vertical
direction c, according to oedometer
tests varies
from a minimum
of 0.06m2/year
to a maximum of 1 m2/year. An estimated
average
is
c, = 0.2m2/year.
The results of the settlement
observations
are presented in Fig. 10.
Excess pore pressure readings were also taken
but the variation in remaining excess pore pressures
(ranging, after nearly 2 years of consolidation,
from
15 to 25 kPa in the middle ofthe clay layer) does not
allow a reliable interpretation.
The effectiveness of
Geodrain and Alidrain is very nearly the same. The
difference in effect between prefabricated
drains
and sand drains with a much larger diameter is also
negligible. Only in the case of 0.9m spacing can
there be found, as expected, a clearly better consolidation effect for the sand drains than for the
prefabricated
drains. A good explanation
of this
result is, as in the case of Sk%Edeby, that the effect
ofsmear is more pronounced in the case of the sand
drains than in the case of the prefabricated drains.
Taking into account the variation in compression
Spacing l-6 m
,
4
,
12
,
16
,
20
,
24
Time: months
. Geodrain
+ Sand drain
xA11drain
Fig. 10. Settlement of ground surface vs time observed at
the &ebro test field
fi
.9
s
G
B
\o
3
.9
and Alidrain.
Grey-brown,
varved
coloured by sulphide
Brown, varved
th. seams silt
The same
Grey-brown,
varved
coloured by sulphide
Grey, slightly
organic
Grey, coloured
by sulphide
The same
Grey, coloured
by sulphide
Grey, varved
th. seams silt
Brown, varved
coloured by sulphide
Grey, slightly
organic
The same
Grey, coloured
by sulphide
Grey, varved
th. seams silt
Brown, varved
coloured by sulphide
Grey, slightly
organic
The same
m*
164
1.62
164
1.53
1.56
1.53
1.71
1.67
1.55
1.55
1.50
97
1.52
1.50
1.45
91
1.53
1.52
61
69
79
82
88
68
106
1.43
1.44
91
60
69
93
115
116
(1)
1.38
1.71
1.63
1.58
1.40
1.32
t-4
54
58
73
90
94
60
93
82
107
102
48
64
61
105
92
(2)
w: y/,
1.26
164
164
1.48
1.42
1.41
(1)
y: t/m3
43
46
54
59
95
46
66
65
94
103
43
53
69
101
104
(1)
36
39
52
66
83
41
65
63
83
113
32
44
45
81
91
(2)
/_
0.25
0.20
0.22
0.28
0.28
0.39
@24
@38
0.41
0.69
0.22
0.24
0.96
0.31
1.2
1.2
1.8
1.4
1.0
1.2
1.0
1.6
0.42
0.42
2.3
1.2
1.3
1.5
1.0
2.8
OCR
(1)
2.0
0.24
0.30
0.31
@30
0.42
0.49
2.5
0.22
0.27
0.39
0.38
0.89
044
0.59
(2)
(1)
tc/%,
I-
of soil in tbe Cirebro test field; groundwater level at about 1 m depth; hydrostatic condition
Grey, slightly
organic
Grey, coloured
by sulphide
The same
Grey, slightly
organic
The same, with
shells
Grey, coloured
by sulphide
Grey, varved
th. seams silt
Brown, varved,
coloured by sulphide
Brown, varved
coloured by sulphide
Grey, slightly
organic
The same, with
shells
Grey, coloured
by sulphide
Grey, varved
* (1) Area with sand drains; (2) areas with Geodrain
t According to fall-cone test.
(I)*
Soil profile
Geological profile and geotecbnical characteristics
Depth:
m
Table 3.
1.0
1-o
1.1
1.5
1.3
1.3
1.3
1.2
1.4
1.1
1.2
2.1
(2)
VERTICAL
Overburden
silty sand
extreme eastern point of the PO river delta,
extensive controlled water loading and other preloading techniques for large steel tanks and many
secondary structures were used (Garassino et al.,
1979). A large number (Z 1700000 m) of vertical
drains of various types (jetted drains, Geodrain,
Soildrain, Sandwick) 27-30m in length were installed with the aim of accelerating the consolidation of a thick stratum of soft young normally
consolidated
(NC) silty clay (cf. Fig. 11). The
geotechnical properties of the clay are presented in
Table 4 (see also Jamiolkowski,
Lancellotta
&
Tordella, 1980). One of the preloading embankments (340m by 65 m) was used as a trial
embankment
with the purpose of comparing the
efficiency of the different drain types installed. In
this Paper the observed behaviour of this embankment (Fig. 12) is reported. The trial embankment
was divided into four sections ofnearly equal length
(about 85m) each one provided with different drain
types. The spacing of the drains was chosen so that
the process of consolidation
according
to the
conventional Terzaghi consolidation theory should
be the same. As a basis for the design the following
values were used: c, = 9m2/year, determined
by
back-analysis
of the behaviour
of a preloading
embankment without vertical drains and representing the upper limit of the laboratory values determined on small-size, good quality, undisturbed
specimens; and c,, = 15 m/year, representing
the
upper limit of the laboratory values obtained from
tests on small-size specimens.
Further in situ testing such as constant head
permeability,
piezometer
probe
dissipation
and preconsolidation
stresses (uvO
, uv,,and
ovc): kN/m*
+ . dvc from 1-D consolidation tests
according to Casagrandes procedure
Fig. 11. Porte Tolle typical soil profile
characteristics
and depth in the test field (and consequently variations in final settlement) a satisfactory correlation
between theory and practice
will be obtained assuming that c,, = 0.8 m/year,
s = 2 and k,/k, = 3 for the sand drains, and s = 2
and k,/k, = 2 for the prefabricated
drains.
Porto
53
DRAINS
Tolle case record
During the construction
of a large (2400 MW)
Thermal Power Plant located in Porto Tolle at the
Table 4. Porto Tolle, soil properties; FVT = field vane test; TC = triaxial compression;
TE = triaxial extension; DSS = direct simple shear; Ck,U = corwolidated k, undrained
Index
properties
Y:
t/m
W:
%
IVL:
0,
,D
Ir:
%
1.85 kO.05
364 f 4.9
52.6 f 6.9
3@9+6.6
Undrained
behaviour
33.9k4.2
E5CJL
C./c
FVT
TC-Ck,U
TE-Ck,U
0.29*
@31
0.19
DE20ur
* Sensitivity
CF:
%
as obtained
from the FVT = 2-3.
DSS-Ck,U
0.26
TC-Ck,U
89
TE-CL&J
82
DSS-Ck,U
87
54
S. HANSBO,
M. JAMIOLKOWSKI
AND
L. KOK
HI
HI
HI
HI
VI-I
n l VI-2
I
Sandwick
Geobrain
I
I
1
h
I
I
I
A
I
I
I
A
I
I
I
A
HI
HI
HI
HI
I
jetted.drain
Soildrain
L,=300
t-
Type of drain
k
Fig. 12
Diameter: mm Spacing:* m
* Triangular
equilateral
Equiva,ent
diameterarray
Plan and cross-section of Porto Tolle experimental
(Ghionna,
Jamiolkowski
& Lancellotta,
1978;
Lacasse, Ladd & Baligh, 1979) and tests with a selfboring pressuremeter (Jezequel & Mieussens, 1975)
permitted an in situ assessment of the c,, leading to
the values given in Table 5. In Fig. 13 the penetration pore pressure measured by means of the
piezometer probe is also reported.
After the embankment was placed, its behaviour
was monitored
by means of 45 piezometers
of
various types, settlement plates, deep settlement
sensors of the bellows-hose type, and vertical and
horizontal
inclinometer
tubes which allowed
measurement
of surface settlement (Fig. 14), vertiTable 5.
Porto Tolle, summary of consolidation
embankment
cal strain, pore pressure (Fig. 15) and horizontal
displacement (Fig. 16).
From Figs 14-16 and from the experience gained
at the site (see also Garassino
et al., 1979) the
performance of the four types of drain used here is
found to be substantially similar in practice. Differences in settlement under different areas cannot be
attributed
to the variations
of soil properties
and/or the thickness of the soft silty clay stratum.
The higher settlement of the area with Sandwick
drains may indicate a greater efficiency of these
drains or a higher disturbance
caused by their
installation than in the case of other types of drain.
properties (from Garassino et al., 1979)
T
Source
1. Laboratory tests*
2. Field permeability plus laboratory compressibility
dissipation tests/piezometer probe
field permeability tests in piezometert
self-boring permeameter
3. Back-analyses
experimental embankment without drains on the basis of pore pressure
preloading embankment with drains on the basis of pore pressure
preloading embankment with drains, from strain and deformation
measurements
*These data available when drainage
t Bilotta & Viggiani
(1975).
system has been designed
Consolidation coefficient
C,:
Ch:
m/year
ml/year
3-9
6-16
22
30
25
19-28
44-66
VERTICAL
Measured penetration
pressure: kPa
nP,
250
500
55
DRAINS
Under the centre of:
Jetted drain area
Geodrain area
Soildrain area
Sandwick area
-.-----
pore
750
0 Elevatron at which
dissipation test has
been performed
6
12
6
1977
1978
Time: number of month (Jan. = 1 etc.)
Fig. 14. Observed surface settlement: Porto Tolle experimental embankment
lo-
=
2
$
g
6
12-
B
3-
12
1977
1978
Time: numbers of month
Fig. 15. Observed excess pore pressure: Porto Tolle
experimental embankment (see also Table 6 for piezometers and additional information.)
18 _
I
20
v
6
22LILL
Fig. 13. Piezometer probe penetration test: Porto Tolle
experimental embankment
12
1977
1978
Time: number of month
Fig. 16. Porto Tolle experimental embankment, maximum (depth- 10m below CL) observed horizontal displacement vs time
S. HANSBO,
56
Undrained
M JAMIOLKOWSKI
shear strength, cu: kPa
40
0
01
80
original F.V. strength
average from 11 19 tests
F.V. strength measured after
embankment
removal
8
t
24
Fig. 17. Porto Tolle experimental embankment,
vane strength before and after preloading
field
Finding the true reason would require a reliable
knowledge of the achieved degree of consolidation
deduced from the pore pressure readings. In this
case such a deduction is not possible because of
problems with the measurement of Au against time.
The back-analysis
of observed
settlements
against time, measured within the soft silty clay
stratum using different velocity methods (Ellstein,
1971, 1972; Scott, 1961), leads to the ch values
shown in Table 5 (for more details see Garassino et
al., 1979), which tends to indicate that when the
embankment load was removed at least 80% of the
consolidation
had been achieved. Such a statement
is supported by the following facts.
(a) Observed vertical settlements are in the
range 75-85% of consolidation
settlements evaluated
using compressibility
properties
back-calculated
from previously available case records of embankments with similar characteristics
of loading and geometry.
(b) The inclinometer
readings indicate only
very small lateral soil displacement during
the period in which large vertical settlement was observed (Fig. 16).
(c) The field vane tests carried out after the
removal of the embankment
show an
increase of undrained shear strength (Fig.
17). This may be explained only on the
basis of the achieved consolidation.
The Au-time trend observed in almost all piezo-
AND
KOK
meters of different types is in contradiction
with
what has been deduced from strain and settlement
observations.
All these piezometers exhibit a very
slow pore pressure dissipation. Those shown in Fig.
15 lead to a consolidation
at points in which
piezometers are located in the range of 2540%.
Such low values cannot be justified even if making
reference to the lowest laboratory-determined
consolidation properties (Table 6). In addition, in many
piezometers, even a very long time after the load
application
when settlement had almost ceased,
significant residual excess pore pressure remains,
showing no tendency to further dissipation.
This case record, representing only a small part
of the large experience gained with vertical drainage problems in Porto Tolle. showed the following.
(4 Vertical drains have performed correctl;.
With the type of preloading
technique
utilized they solved, in an economical way,
many important foundation problems.
speed de(b) The observed consolidation
duced from the settlement observations
is
appreciably higher than that which can be
predicted on the basis of the consolidation
properties determined by means of in situ
tests given in Table 5. This may be explained by the trace of the penetration
pore pressure curve in Fig. 13, which
indicates
frequent
occurrence
of thin
coarse seams (probably
silt horizons).
When these seams are interconnected
by
vertical drains, there is on average a very
short vertical drainage path. The presence
of seams may thus make a substantial
contribution
to the consolidation
by
vertical flow.
(c) The delayed excess pore pressure dissipation, or even lack of it (with the trapped
values of Au, 30_40kPa,
in excess of
existing static pore pressure), which is in
contrast with all other information gained
from the monitoring programme, cannot
be explained at the present stage of the
Authors knowledge.
Tries&
case
record
In early 1969 an area of about 1.5 km*, reserved
for the extension of an existing refinery, was filled
up under uncontrolled
geotechnical
conditions
with regard to fill thickness and relevant soil
The discussed piezometer performance is typical for the
majority of about 150 piezometers
of different types
installed in Porte Tolle on areas with vertical drains.
z On the other areas with vertical drains few piezometers
showed a reasonable AU dissipation leading to the backcalculated c,, value given in Table 5 and in agreement with
other in situ tests.
VERTICAL
Q
HE=36m
HF= 45-5-O m
R=375m
! Preloading embankment
\ placed August 1973
Fill placed 1969-l
57
DRAINS
I
\
sr
970
4.5 m + MSL
F= 0.0
Very soft to soft silty clay containing shell fragments
and uniformly distributed traces of organic matter
5
;
NVI = northern vertical Inclinometer
SW = southern vertical inclinometer
tube
tube
May 1973: 100 mm diameter, 25 m long SandwIcks
mstalled in equilateral triangular array, spacing 1.9 m
125
Fig. 18. Trieste preloading embankment,
cross-section
characteristics
(Fig. 18). Extensive in situ and
laboratory
investigations
were later carried out,
leading to the geotechnical characterization
of the
soft NC clay stratum beneath the fill, summarized
in Table 7.
In March 1973, after it had been decided to build
two 50000 m3 crude oil steel tanks on the examined
area, displacement-type
sand drains (Sandwicks),
0.10 m in diameter, were installed with 2 m spacing
in an equilateral triangular pattern at the sites of
the tanks to be built. The drainage system was
designed,
assuming
c, = 0.6 m*/year
and
ch = 1.2 m/year, representing the upper limit of the
laboratory
values determined
on good quality,
small-volume,
undisturbed
specimens (see Fig. 19
and Jamiolkowski,
1974).
Table 6.
Porte Tolle, piezometer
In June 1973, two embankments, 4.0 m and 3.6 m
in height, were placed on the drained area, representing in the intention of the designer a first lift of a
two-stage preloading embankment
with the aim of
anticipating
the consolidation
of the soft clay
stratum before the construction
of the steel tanks.
The lower embankment
(3.6m) which is considered here was instrumented as shown in Fig. 20.
The pore pressure measurements,
made before the
placement of the fill, revealed that in June 1973 the
soft clay stratum was practically unconsolidated
under the weight of the general fill dumped over the
area in 1969.
During and immediately after the placement of
the first preloading lift, large vertical settlement and
horizontal displacement (Fig. 21) took place within
types used, as shown in Fig. 15
Observed
Predicted1
W%,,,
W~umax
Time in days
Type of
piezometer
+ Hydraulic, Bishop type
0 Electra-pneumatic
x Vibrating wire, Maihak
0 Hydraulic, Bishop type
AElectrical, BAT type
VElectrical, BAT type
Depth: *
m
24.0
19.4
19.3
24.0
12.2
20.2
Drain type/
areat
Jetted drain
Soildrain
Geodrain
Sandwick
Geodrain
Jetted drain
*Below ground level at depth 0.10 m below MSL.
t Under the centre of each area and halfway from the drains.
:Terzaghi
consolidation
theory with no allowance
for smear
consolidation
properties: c, = c,, = 9 m2/year.
and
175
345
175
345
0.807
0.812
0.893
0.824
0.718
0.893
0.594
0,673
0.767
0.577
0.565
0,699
0.510
0.534
0.534
0.520
0.534
0.538
0.260
0.290
0,290
0.276
0.290
0.293
minimum
possible
well resistance
assuming
58
S. HANSBO.
M. JAMIOLKOWSKI
the soft clay stratum, with effectively no excess pore
pressure (Au) dissipation recorded, indicating pronounced confined plastic flow within the clay
formation
(estimated
safety factor against undrained bearing capacity failure NN1.15-1.30).
In the following 9 months large vertical settlement and increasing lateral displacements
were
observed, together with very slow pore pressure
Coefficient
2x10-1
0
.
E
of consolidation,
I
.
0.
KOK
decay (Fig. 22). This decay is in accordance with the
conventional Terzaghi consolidation
theory adopting c, = c,, = 0.3m2/year
corresponding
to the
lower limit of the laboratory
determined
values
of c,.
In April 1974, because of the observed very slow
rate of consolidation,
the idea of using the preloading technique was abandoned
and the embank-
cy, ch: m/year
10
20
1
.
A
*
AND
.
.
.*
+
.
.
.
18L
Vertical flow. C,:
Horizontal flow, c,,:
l incremental
loading oedometer
A controlled hydraulic gradient
oedometer
x triaxial dissipation test
--
laboratory
Fig. 20.
tion
Trieste preloading embankment,
Trieste, soil properties
Index
properties
Y:
t/m3
w:
%
1.72 _+0.04
49.9 f 4.0
Undrained
behaviour
WL,:
%
70.7* 11.5
I,:
%
CF:
%
47.4 & 10.0
45.8 & 6.7
OJls
FVT
TCXk,U
0.35*
0.32
Drained
behaviour
*Sensitivity
settlement stack
pneumatic warlam-type piezometer
brs mercury settlement gauge
vertical inclinometer tube
horizontal inclinometer tube
0 incremental loading oedometer
+ triaxial dissipation test
Fig. 19. Trieste preloading embankment,
determined coefficients of consolidation
Table 7.
TE-Ck,U
DSS-Ck,U
TCXk,U
0.28
0.26
TCCk,U
as obtained
Eu591GP
Oedometer
90
tests
I$?
degrees
c:
t/m2
CR
c,:
m/year
26
0.24
0.34.9
from the FVT = 24.
TE-Ck,U
77
instrumenta-
VERTICAL
59
DRAINS
160-
:5
lR
.
8 .M.
measured
x theoretical
a
: 120-
x**m.
.xA
.
.-.
S, at 10 m below GL
12
1973
1974
Time: numberofmonth
Horizontal inclinometer tube,
i.e. vertical displacement under the
centreline of embankment
6
A Northern vertical inclinometer tubes
. Southern ? maximum horizontal displacement
x Vertical settlement under the embankment
centre caused by lateral soil displacement
Fig. 21. Trieste preloading embankment,
cal and horizontal displacements
observed verti-
ments were therefore removed. In 1975, two floating roof tanks were constructed
on the drained
areas. These tanks were subjected to storage restrictions, controlled by the measurements ofconsolidation of the soft clay stratum.
Finally, in November
1978, it was possible to
allow unrestricted
use of the two tanks under
examination. In July 1980 the observed settlement
of the tanks, except that which occurred under the
preloading
embankment,
reached
values
of
1.2-1.4 m under the centre and 08-1.0 m under the
perimeter, with the primary consolidation
process
almost totally completed.
During the period 1975-80 (Ghionna et al., 1980)
additional in situ and laboratory soil investigations
were carried out with the aim of understanding
better the geotechnical behaviour of the Trieste clay
and obtaining a deeper insight into the examined
case record. The investigations
using the piezometer probe and the self-boring
pressuremeter
obtained the following information relevant to the
vertical drainage problem.
pore pressure, not reported
(a) Penetration
here, revealed a complete lack of any more
permeable layers of lenses within the clay
mass.
tests with the piezometer
(b) Dissipation
probe indicated a range of c, from 1.2 to
1.4 m/year.
pore pressure deduced
(c) The equilibrium
Time: number of month
Fig. 22. Trieste preloading embankment,
meter record
typical piezo-
from piezometer
probe dissipation
tests
and from relaxation tests run with the selfboring device gave as a result the July 1980
distribution
of Au beneath
the area
without drains that is shown in Fig. 23.
This confirms very low clay consolidation
properties with respect to the vertical flow.
The examination of the Trieste case record allows
the following comments.
(4 The vertical drainage design and performance must be considered
as unsatisfactory. The consolidation
process unexpectedly coincides with the lower limit
values c, = ch = 0.3 m/year determined
in the laboratory
on small-size undisturbed clay specimens.
(b) In respect to this, why, at the considered
site, is the field consolidation
so slow?
What is the influence on the consolidation
process of the measured large, virtually
undrained,
lateral and vertical displacements?
The change in soil structure that would follow upon
large undrained deformations
may lead to a reduction of the coefficient of consolidation;
and the
conventional
consolidation
theory may be inadequate when applied to high plasticity clays with
pronounced
viscous response, particularly at high
stress level. These theories may at least partially
explain the peculiar behaviour
of the Trieste
embankment.
AziShaven case record
For quick and efficient installation of prefabricated drains, contractors
offer in principle two
methods.
In the static type of installation a heavy machine
60
S. HANSBO,
Effective (d,J
Fig. 23.
M. JAMIOLKOWSKI
and total (u,J overburden
Trieste preloading embankment,
AND
L. KOK
stress and pore pressure (14: kPa
overburden stresses and pore pressure in the area without vertical drains
Table 8. Characteristics of the installation
(Azighaven and Hemspoor case records)
Industrial site
Characteristics
Depth contour
Penetration velocity
(mandrel)
Frequency (vibrator)
Cross-sectional area
of mandrel
t
Aziehaven(harbour)
Fig. 24. Azifhaven,
clayipeat layer
location of test area and depth of
with sufficient thrust forces a mandrel down to the
required depth. This machinery has a high production rate, but has to be assembled on the site.
In the dynamic type of installation the mandrel is
either hammered down (as in piling) or the driving
force is delivered by a vibrating hammer. These
vibrating hammers, with working frequencies of
Static
z 0.5m/s
52110cm
techniques
Dynamic
zlm/s
25 Hz
50 cm2
8-50 Hz, can be electrically or electro-hydraulically
operated. As the vibrating mandrel eliminates an
important
part of the friction along the shaft, a
light crane can handle the combination of vibrating
hammer, guidemast and mandrel (Table 8).
To study the effect of the two installation techniques on pore pressure and consolidation
response, a new industrial site in the western harbour
area of Amsterdam was chosen. The actual site was
intersected by an old gully, which was completely
filled with naturally sedimented layers of clay and
peat. The geographical layout of the area and some
depth contour lines are given in Fig. 24. This area
needed improvement
as far as soil conditions were
concerned. It was decided to accelerate the consolidation of the compressible layers by means of
vertical drains and preloading.
In view of the techniques the area was divided
into two sub-areas. In each sub-area 150000m of
prefabicated drain (Geodrain) were installed, static-
VERTICAL
61
DRAINS
Cone resistance: MPa
10
Bulk density:
20
m3
Clayey
peat
Clay
slightly
organic
1.4
area
Fig. 25.
ACehaven, schematic view of test area
ally as well as dynamically. A schematic view is
given in Fig. 25.
Additional subdivision was provided to detect
differences in the drain spacing. Triangular grids
with drain spacings of 1.2 and 1.8 m were used. The
characteristics
of the soil strata encountered
are
shown in Fig. 26.
Direct excess porewater pressure response. Electrical pore pressure devices were installed beforehand
in the anticipated triangular grids. A general picture of the excess pore pressure behaviour during
and after installation of the prefabricated
drains is
given in Fig. 27. These results were obtained from
the clay layer at depths of 14-15 m below datum. In
the static case, relatively small excess and underpressure with respect to hydrostatic pore pressure is
generated. In the dynamic case the pore pressure is
significantly higher. If it is assumed that the porewater pressure generation in the static case can be
calculated according to the theory of an expanding
cavity, an estimate can be made of the relation
between excess porewater pressure and distance
from the mandrel. The result of this calculation,
made according to VesiC (1972), is indicated in Fig.
28. The magnitude of the calculated pore pressure is
dependent on the unconfined eompressive strength
(cohesion), but independent
of the cross-sectional
area of the mandrel.
The pore pressure distribution
in soil space is
dependent
on the cross-sectional
area of the
Fig. 26.
Azilhaven, soil characteristics
Piezometer
(14-l 5 m below DL)
,
Prefabricated
drain
Time: hours
Fig. 27. Azikhaven, excess pore pressure after static and
dynamic installation
mandrel. If this is so then high pore pressures are set
up in the direct neighbourhood
of the mandrel
during penetration,
and in the plastic zone very
close to the mandrel the generated excess pore
pressure may exceed the original effective stress
62
S. HANSBO, M. JAMIOLKOWSKI
AND L. KOK
Time: days
160
Effective stress
Fig. 29.
04
Distance, R: m
320
\
Prefabricated
drain
l -e
v-v
O--O
02
240
06
Fig. 28. Azkhaven, excess pore pressure vs distance from
the mandrel
(Au/&>l).
This criterion may be used to determine a zone of disturbance
(remoulding) around
the mandrel. Arbitrarily,
Au/& = 0.8 might be
chosen as a criterion for the extension of the zone of
smear. From the pore pressure coefficient the
distance R can be determined and chosen as d, in
the dimensionless
form d,/d = s. This criterion
would be invalidated
if permeable
layers in
stratified soils are blocked horizontally
due to a
dragdown of clay by the mandrel. In certain soils
mandrel penetration can cause additional fracturing of the clay layer and improve the permeability
of the surrounding
soil of the installed prefabricated drain. A substantial
effect on drain performance can be achieved if the cross-sectional
areas of the prefabricated
drain and mandrel,
including the anchor plate, are well adapted to the
characteristics of the soil to be consolidated. Where
prefabricated
drains are used to improve
the
stability of soil structures account must be taken of
the generated excess porewater
pressure during
installation. For that purpose static installation is
to be preferred.
Direct consolidation response. The sub-areas were
equipped with instruments to record various events
like pore pressure, settlements and horizontal displacement. From the settlement records in both
areas the consolidation
velocity was calculated
with respect to time. No distinction is made here
with regard to drain spacing (1.2 and 1.8 m). The
excess porewater
pressure was not taken into
no drains
static
dynamic
drainspacing
1.2 m
Azkhaven, rate of consolidation
account in the presentation
of the consolidation
process because of the inexplicably high excess pore
pressures compared with recorded settlement. No
significant difference between the two methods of
installation can be made with regard to the rate of
consolidation
(Fig. 29).
Hemspoor case record
This record shows a pseudo-consolidation
effect
with respect to piercing the soil layers to be
consolidated.
A new elevated railway track had to be built very
close to the city of Amsterdam and the alignment of
the embankment
had to cross an old gully, naturally filled with very soft clay (Fig. 30). The depth of
this very compressible layer is approximately
25 m.
The design of the elevated embankment
shows a
maximum height of 5 m above datum and it was
assumed that about 13m of sand was needed to
reach this design height (Fig. 30).
To obtain sufficient stability during the construction, and a minimum of end settlement in the long
term, vertical drainage by means of prefabricated
drains was envisaged. Prefabricated
drains were
chosen because of successful applications in other
projects (Aziehaven case record) and the hazards to
be expected during the installation of sand drains
over 25 m length.
Characteristics
of the soil, which consists of soft
clay, except just below original ground level, are
shown in Fig. 31. Prefabricated
drains were installed in a triangular grid with spacing of 1.2 m by
dynamic means. Piezometers
were installed at
various depths in the soft clay and the first fill was
placed. The filling operations and the appropriate
pore pressure readings can be seen in Figs 32 and
33. When the prefabricated
Geodrains
were installed, an immediate pore pressure response was
obtained.
The excess pore pressure
dropped
dramatically, but there was concern when a new
load (dry sand fill) was placed and there was no sign
VERTICAL
Fig. 30.
63
DRAINS
Hemspoor, view and cross-set tion of test area with depth of soft layer
of a quick release of excess porewater pressure.
Additional prefabricated
drains were installed to
obtain a spacing of 0.6m. The excess porewater
pressure responded immediately. Again filling the
site resulted in a load-dependent
increase in pore
pressure, without a significant drainage effect. The
results are given schematically in Fig. 34.
No satisfactory
answer for, this odd drainage
behaviour has been given yet. This effect could be
initiated by the installation of the drain itself The
large cross-sectional
area ratio between mandrel
and prefabricated
drain may cause the release of
excess pore pressure. The pore pressure coefficient
r, recorded during the normal filling procedures
with dry sand, varied between 0.6 and 0.7. An
exceptional decrease of the pore pressure coefficient
was observed directly and for three months after
installing the prefabricated
drains.
CONCLUSIONS
Installation
AND
drains; this release has no direct connection
consolidation
by drain wells in itself.
Cone
resistance:
MPa
Soft clay
RECOMMENDATIONS
effects
There is a remarkable difference in the excess
porewater pressure generation between the static
and the dynamic installation techniques.
There is no significant
difference, it seems,
between dynamic and static installation techniques
in respect of consolidation.
A remarkable
excess pore pressure release is
recorded
immediately
after the installation
of
25
Fig. 31.
Hemspoor, soil characteristics
Bulk
density:
1.35
with
64
S. HANSBO,
M. JAMIOLKOWSKI
AND
L. KOK
Time
Fig. 32.
Hemspoor, pore pressure and fill height vs time
.~_~.____
NAL
Groundwater
Original ground
_~
table
level
1977
1978
1979
Time
Fig. 33.
Hemspoor, depth of original ground level and fill height vs time
When prefabricated
drains are used to improve
the stability of, for instance, slopes of foundations,
the dynamic
installation
technique
should be
avoided. Otherwise the installation of drains may
occasionally affect the existing stability condition.
Excess pore pressures
are generated
by the
insertion of the mandrel. This can lead to an
extended zone of disturbance.
The ratio of crosssectional area of drain and mandrel (anchor plate)
should be carefully adapted
to the soil. More
attention should therefore be given to the installation equipment.
The adopted
installation
procedure and related modifications
of properties of
the soil surrounding
the drain have a great
influence on the drainage system performance and
therefore this factor has to be considered in the
design. The fact that drains with different diameters
but equal spacing may give the same consolidation
effects can be explained by different values of s or
&A.
Process of consolidation
Notwithstanding
good progress made in laboratory and field techniques allowing determination
of
k, or c,,, the reliable assessment of these critical
design parameters remains one of the most difficult
tasks within the area of geotechnical engineering.
Even if in the majority of the examined case records
c,,&c,, an uncritical and generalized assumption
that this is always the case is not justified. For long
prefabricated
drains and small-diameter
sand
drains, well resistance has to be considered in the
VERTICAL
Datum line
DRAINS
65
p,j
I
1977
I
1978
J
1979
Time
Fig. 34.
Hemspoor,
schematic result of excess pore pressure and surcharge vs time
design of the drainage system. Well resistance can
explain why the relative increase in shear strength
due to consolidation
gets smaller with depth.
In many cases the rate of consolidation estimated
from vertical settlement observations is higher than
that estimated from excess pore pressure dissipation, although almost no lateral movements have
occurred. The fact that an increase in undrained
shear strength has been observed in spite of an
almost unchanged excess pore pressure is a reason
to rely more on settlement than on pore pressure
dissipation. Since in some cases excess pore pressure remains although further long-term settlement
is negligible, the Authors recommend
the use of
settlement as a measure of the process of consolidation.
Experienced
engineering
judgement
is very
necessary with respect to several uncertainties
in
parameter determination,
design and drain performance, to avoid disappointing
results.
REFERENCES
Akagi, T. (1977). Effect ofdisplacement type sand drains on
strength and compressibility qf sqft clays. Dept Civ.
Engng, Tokyo University.
Barron, R. A. (1948). Consolidation
of fine-grained soils
by drain wells. Trans. Am. Sor. Civ. Engrs 113, Paper
No. 2346.
Billotta, E. & Viggiani, C. (1975). Unindagine
sperimentale in vera grandezza sul comportamento
di un
banco di argille normalmente
consolidate.
12th Ital.
Conf: Soil Mech., Cosenza 1, 223-240.
Ellstein, A. (1971). Settlement prediction
through
the
sinking rate. Rev. Lat. Am. Geotec. 1, No. 3, 231-237.
Ellstein, A. (1972). Settlements development: observed vs.
predicted. Purdue conference on performance of earth
and earth-supported structures, Vol. 1, Part
2,
1073-1085. New York: American
Society of Civil
Engineers.
Garassino,
A., Jamiolkowski,
M., Lancellotta,
R. &
Tonghini,
M. (1979). Behaviour
of pre-loading
embankments
on different vertical drains with reference to soil consolidation
characteristics.
7th Europ.
Co@ Soil Mech, BrightoIl 3, 213-218.
Ghionna, V., Jamiolkowski,
M. & Lancellotta,
R. (1978).
Determinazione
de1 coefficiente di consolidazione
da
prove di disdpazione
eseguite con la punta piezometrica.
13th Ital. Conf: Soil Mech. Fdn Engng.
Merano, Italy, discussion.
Ghionna,
V., Jamiolkowski,
M., Lancellotta,
R. &
Tordella, M. (1980). Performance of self-boring pressuremeter in cohesice deposits. Policnico di Torino,
Italy (in preparation).
Hagerty, D. J. & Garlanger,
J. E. (1972). Consolidation
effects around
driven piles. Purdue conference on
performance of earth and earth-supported structures,
Vol. 1, Part 2, 1207-1222.
New York: American
Society of Civil Engineers.
Hansbo, S. (1960). Consolidation
of clay with special
reference to influence of vertical sand drains. Proc.
Swed. Geotech. Instn. No. 18.
Hansbo, S. (1979). Consolidation
of clay by band-shaped
prefabricated
drains. Ground Engng 12, No. 5, 1625.
Hansbo, S. (I 98 1). Consolidation
of fine-grdined soils by
prefabricated
drains. Proc. 10th Int. Conf Soil Mech.,
Stockholm (to be published).
Holtz, R. & Broms, B. (1972). Long-term loading tests at
Ska-Edeby,
Sweden. Purdw con@rence on performance of earth and earth-supported structures. New
York: American Society of Civil Engineers.
Jamiolkowski,
M. (1974). Proceedings of conference on
66
S. HANSBO,
M. JAMIOLKOWSKI
settlement of structures, Cambridge, discussion, 6W
694.
Jamiolkowski,
M., Lancellotta,
R. & Tordella,
M. L.
(1980) Geotechnical
properties
of Porto Tolle NC
silty clay. 6th Danube Europ. Conf: Soil Mech., Varna,
Bulgaria.
Jezequel, J. F. & Mieussens, C. (1975). In situ measurement of coefficients of permeability and consolidation
in fine soils. Proceedings of conference on in situ
measurement of soil properties, Raleigh, Vol. I,
208-224.
New York: American
Society of Civil
Engineers.
Kok, L. (1978). Explosion
densification
and vertical
drainage as soil consolidation
techniques for harbour
AND L. KOK
building. 7th Inc. Harbour Con& Antwerp.
Lacasse, S. M., Ladd, C. C. & Baligh, M. M. (1978).
Evaluation offeld vane, Dutch cone penetrometer and
piezometer testing devices. Res. Rept, Dept Civ. Engng,
MIT, Cambridge,
U.S.A.
Scott, R. F. (1961). New method of consolidation
coefficient evaluation.
J. Soil Mech. Fdns Div. Am. Sot.
Civ. Engrs 87, SMl, 29-39.
Vesit, A. S. (1972). Expansion of cavities in infinite soil
mass. J. Soil Mech. Fdns Div. Am. Sot. Civ. Enars.
SM3, 265-290.
Yoshikuni. H. & Nakanodo.
H. (1974). Consolidation
of
fine-grained
soils by drain wells with finite permeability. Jap. Sot. Soil Mech. Fdn Engn 14, No. 2.
