Consolidation beneath Circular Skirted Foundations

Susan Gourvenec1 and Mark. F. Randolph2

Abstract: The effect of foundation embedment on consolidation has not been considered previously in a systematic manner, although this
is of particular interest for offshore foundations, where embedment is provided by skirts that enclose a compressible soil plug. For skirted
foundations, critical uncertainties include what to assume in terms of the degree of drainage at skirt tip level, and the relative time scales
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

of consolidation within the soil plug and beneath the foundation. In this paper, results from small strain finite-element analyses are used
to quantify the immediate and time-dependent response of circular skirted foundations to uniaxial vertical loading. Foundations with
frictionless and fully rough skirt-soil interfaces with varying ratio of embedment depth to foundation diameter are considered and the
responses compared with those for surface foundations. The findings show that both skirt-soil interface roughness and embedment ratio
have a significant effect on the consolidation response.
DOI: 10.1061/共ASCE兲1532-3641共2010兲10:1共22兲
CE Database subject headings: Soil consolidation; Foundations; Offshore structures.

Introduction dissipate compared with a conventionally embedded foundation.

Skirted foundations are used widely offshore, either as a single
A theory for three-dimensional consolidation was first formulated foundation system for gravity based structures, as discrete foun-
by Biot 共1935, 1956兲. A number of analytical solutions of the dation units at the corners of jacket structures and tension leg
time-settlement behavior of shallow foundations have since been platforms, or as foundations for subsea infrastructure such as
proposed for a range of permeable to impermeable and flexible to manifolds. Some applications of skirted foundations offshore are
rigid rafts, typically assuming a smooth interface and an underly- illustrated in Fig. 2. Foundation footprints for large gravity based
ing elastic half space 关e.g., McNamee and Gibson 共1960兲, Gibson structures can have an equivalent diameter of up to 100 m and
et al. 共1970兲, Booker 共1974兲, Chiarella and Booker 共1975兲, and skirt depth to foundation diameter ratios 共d / D兲 typically in the
Booker and Small 1986兲兴. Little attention has been directed to- range 0.1ⱕ d / D ⬍ 0.4 关e.g., Tjelta et al. 共1990兲, Hansen et al.
ward quantifying the effect of foundation embedment on consoli- 共1992兲, Humpheson 共1998兲, and Watson and Humpheson 共2007兲兴,
dation response. An analytical solution to the problem of an individual skirted foundations for jackets or tension leg platforms
embedded foundation would involve considerable computational have diameters between 10 and 20 m with embedment ratios in
difficulty and to the writers’ knowledge the problem has not been the range 0.3ⱕ d / D ⬍ 1.0 关e.g., Støve et al. 共1992兲 and Bye et al.
previously addressed numerically in a systematic manner. 共1995兲兴, while small foundations for subsea installations might be
Embedment would be expected to reduce the magnitude and only a few meters in plan dimension with embedment ratios typi-
rate of consolidation settlements of shallow foundations due to cally less than 0.1. In general, skirted foundations with embed-
the resistance of the soil above foundation level and longer drain- ment ratios less than unity are employed in the field, particularly
age paths. Additionally, for rough sided foundations consolidation for foundations with a large footprint for which higher embed-
settlements will be reduced due to a portion of the applied load ment ratios represent considerable embedded depth. Higher as-
being carried by side friction. For offshore applications, embed- pect ratio “suction caissons” are used as anchors for mooring
ment generally takes the form of skirted foundations, where a lines, but there the loading is generally applied at a significant
plug of soil is confined beneath the foundation by circumferential angle away from the vertical. Suction caissons are outside the
skirts, as illustrated schematically in Fig. 1. The additional in- scope of what is considered here, although the effect of sustained
crease in drainage path length compared to a conventionally em- uplift loading on capacity has been considered by Clukey et al.
bedded foundation 共from under the base plate to skirt tip level and 共2004兲.
then to the free surface兲 and additional compression of the con- Understanding the consolidation response of skirted founda-
fined soil plug will lead to larger consolidation settlements and tions is necessary to predict the rate and magnitude of settlement
longer times for the excess pore pressures beneath the cap to for foundations subjected to compression. Another interest in the
consolidation response of skirted foundations exists for applica-
1
Associate Professor, Centre for Offshore Foundation Systems, The tions that impose tensile loads, as transient suctions developed
Univ. of Western Australian, 35 Stirling Hwy, Crawley, Perth, WA 6009 within the soil plug enhance the uplift resistance. While passive
共corresponding author兲. E-mail: [email protected] suctions can be maintained, a skirted foundation may mobilize
2
Winthrop Professor, Centre for Offshore Foundation Systems, The uplift resistance from “reverse” end bearing. With time water
Univ. of Western Australian, 35 Stirling Hwy, Crawley, Perth, WA 6009.
flows into the soil plug, the suctions dissipate and the uplift ca-
Note. This manuscript was submitted on February 6, 2008; approved
on September 1, 2009; published online on January 15, 2010. Discussion
pacity is eventually reduced to that provided by the weight of the
period open until July 1, 2010; separate discussions must be submitted for foundation and the friction mobilized along the skirts. The time
individual papers. This paper is part of the International Journal of scale over which suctions can be relied upon thus becomes criti-
Geomechanics, Vol. 10, No. 1, February 1, 2010. ©ASCE, ISSN 1532- cal in design. This study concentrated on the consolidation
3641/2010/1-22–29/$25.00. response of skirted foundations nominally subjected to compres-

22 / INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010

Int. J. Geomech. 2010.10:22-29.

 A D Geometry and Material Parameters
Circular foundations with skirt depth to diameter embedment ra-
B B d
tios d / D = 0 共surface plate兲, 0.15, 0.3, 0.5, and 1.0 and skirt thick-
A
t ness t / D = 0.002 共based on typical field conditions兲 were
Section A - A
represented in axial symmetry. The soil was represented by a
D homogeneous elastic half-space with Biot-type consolidation gov-
Foundation diameter D erning the stress-pore fluid coupling. Constant and isotropic elas-
Embedment depth d
Skirt thickness t tic parameters Young’s modulus E⬘ and Poisson’s ratio ␯⬘ were
t
assumed in all analyses, but isotropic and anisotropic permeabil-
ity were considered with ratios of horizontal to vertical perme-
ability kh / kv = 1, 3, and 10. Results are presented in terms of
Section B - B dimensionless quantities 共d / D, cvt / D2, and wE⬘ / D⌬q兲, and are
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

therefore independent of the actual values of foundation diameter,

Fig. 1. Schematic representation of a skirted foundation embedment depth, soil stiffness, or permeability adopted in the
finite-element model. As such, the results are equally applicable
to any foundation diameter, embedment depth, soil stiffness and
permeability 共provided consistent units are observed兲.
The magnitude of Poisson’s ratio affects the three-dimensional
consolidation response with higher Poisson’s ratio leading to
faster consolidation and a reduction in the ratio of consolidation
settlement to total settlement. However, over a practical range of
drained Poisson’s ratio 0.1ⱕ ␯⬘ ⱕ 0.3, the effect is limited 共Schiff-
mann et al. 1969; Booker and Small 1986兲. A typical value of
drained Poisson’s ratio ␯⬘ = 0.2 was selected for the analyses in
this study.
Interaction between the foundation and soil was represented
by either a fully rough or frictionless interface to encompass the
possible range of interface friction of different foundation mate-
Gravity base Jacket Tension leg Subsea
rials and finishes.
structure platform structure
(GBS) (TLP)
Finite-Element Mesh
Fig. 2. Applications of skirted foundations offshore Each foundation was represented as a rigid body and the soil as a
deformable solid with first order fully integrated axisymmetric
sion although the effect of embedment ratio and skirt-soil inter- stress-pore fluid continuum elements. The foundation was in
face roughness on the underlying mechanisms governing excess place at the start of the analyses 共i.e., installation was not mod-
pore pressure dissipation would be similar for foundations acting eled兲. Zero-displacement boundary conditions were prescribed in
in tension within the limits of the elastic soil response assumed the radial direction around the circumference and vertically across
here. the base of the mesh, located sufficiently remote so as not to
This study has considered the consolidation response beneath affect the foundation or consolidation response. During consoli-
impermeable, rigid, skirted foundations with skirt depth to foun- dation, a zero pore pressure boundary condition was prescribed
dation diameter ratios d / D of 0.15, 0.3, 0.5, and 1.0 共to represent along the surface of the mesh representing the soil while the
the range of “shallow” foundations兲, for both smooth and rough foundation was considered impermeable. Drainage was not per-
skirt-soil interfaces, embedded in an elastic soil. Results are pre- mitted across the circumferential boundary of the mesh.
sented in terms of the initial contact pressure and stress distribu- Each of the meshes was constructed to a similar template,
tion in the soil, excess pore pressure dissipation and the time- Fig. 3 shows the mesh for an embedment ratio of d / D = 0.5 as an
settlement response. Results from the analyses of skirted example. Each mesh maintained a uniform element size along the
foundations are compared with numerical and analytical solutions skirts l / D = 0.025, where l defines the element length along the
for a surface foundation 共d / D = 0兲. This study considered ideal- skirt-soil interface.
ized elastic conditions to identify the principal mechanisms gov- Fig. 4 illustrates the mesh dependency for the smooth skirted
erning the consolidation response of skirted foundations as a foundation with embedment ratio d / D = 0.5, in terms of the dis-
function of embedment ratio and interface roughness and to en- tribution of contact pressure beneath the base plate, taken as the
able comparison with established analytical solutions for a sur- initial excess pore pressure ⌬ui normalized by the applied foun-
face foundation. The effect of soil yielding on the consolidation dation pressure ⌬q. Three mesh configurations were considered,
response of skirted foundations would provide a useful extension each with uniform elements along the foundation skirts, a “coarse
to the set of benchmark solutions identified in this study. mesh” l / D = 0.05, a “fine mesh” l / D = 0.025 共as adopted in the
study兲, and a “very fine mesh” l / D = 0.015. 共Other configurations
Finite-Element Model were investigated, for example with finer mesh discretization con-
centrated at skirt tip level, but these were found to be more mesh
Small strain finite-element analyses were carried out with the sensitive.兲 Increasing mesh density attracts a greater proportion of
commercially available software ABAQUS 共Dassault Systèmes the applied load to the skirt tips, with less being carried by the
2008兲. base plate. The initial contact pressure at the mid point of the

INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010 / 23

Int. J. Geomech. 2010.10:22-29.

 tions prevailed in the analyses. In the second step, a drainage
x
boundary was specified along the soil surface 共either side of the
D/2 foundation兲 and consolidation was permitted until excess pore
pressures had dissipated.
Axis of symmetry

An important issue in consolidation problems is the choice of

d the initial time step. As the governing equations are parabolic the
initial solution immediately after a change in boundary condition
is a local solution. Time steps smaller than a certain duration give
meaningless results and distort the subsequent consolidation re-
sponse. The time step over which the excess pore pressure regime
was set up and the minimum time step permitted during consoli-
dation was defined by 共Vermeer and Verruijt 1981兲
z
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

Fig. 3. Finite-element mesh for d / D = 0.5. 关Note: the foundation is in ␥w

tmin = h2 共1兲
place at the start of the analyses. The foundation skirts are visible as 6Ek
a heavier line at 共x , z兲 = 共D / 2 , 0兲 to 共D / 2 , d兲. A single column of
elements is provided beneath tip level, of equal width of the skirt tips where h = characteristic element size 共i.e., the distance between
t / D = 0.002, but are too thin to be seen at figure scale.兴 the Gauss points兲 near the draining surface; ␥w = unit weight of the
pore fluid; E = elastic modulus of the soil skeleton; and k = soil
permeability.
1.2

t/D = 0.002 Benchmarking

1.0

The finite-element model of a smooth surface foundation was

0.8
benchmarked against available analytical solutions for initial con-
tact pressure 共Muki 1961兲 and time-settlement response 共Booker
∆ui/∆q

and Small 1986兲. Fig. 5 共and later Fig. 10兲 shows good agreement
0.6
t/D = 0.02 x

0.4
D/2 was achieved between the numerical and analytical solutions.
Coarse mesh, l/D = 0.05 The established analytical studies cited here and the numerical
d Fine mesh, l/D = 0.025
0.2 study presented in this paper represent the soil mass as an elastic
Very fine mesh, l/D = 0.015
t half-space. Therefore, the ensuing results should be considered in
0.0 the context of loading situations appropriate to assumptions of
0.0 0.1 0.2 0.3 0.4 0.5
x/D
elasticity, particularly that loads are of magnitudes less than
would cause significant local yield within the soil mass.
Fig. 4. Illustration of mesh dependency in terms of initial excess
pore pressure distribution beneath foundation base plate for smooth
sided foundation with d / D = 0.5 Results

foundation with a skirt thickness ratio t / D = 0.002 共as adopted in Initial Excess Pore Pressure Distribution
this study兲 ranges between 0.92⬍ ⌬ui / ⌬q ⬍ 0.98 depending on Fig. 5 shows the initial excess pore pressure ratio ⌬ui / ⌬q along
mesh density. Comparison with a skirt thickness ratio t / D = 0.02, the interface with the foundation base plate immediately follow-
shows increasing mesh dependency with increasing skirt thick-
ness. As would be expected, thicker skirts attract more load to
skirt tip level, leaving less load carried by the base plate. The
higher stress concentration at skirt tip level leads to a greater Smooth d/D Rough x

D/2
mesh dependency with the initial contact pressure at the mid point 0
1.4
of foundation ranging between 0.60⬍ ⌬ui / ⌬q ⬍ 0.88 for the mesh
0.15
Muki (1961)
0.3 d
densities considered. Only the smooth foundations exhibited 1.2 0.5

mesh dependency. The response of the rough foundations was 1

1
mesh independent due to portion of load carried by skirt friction
leading to lower stress concentrations at tip level. 0.8
∆ui/∆q

0.6
Scope and Loading Method
0.4
Each analysis was carried out in two stages. In the first stage,
excess pore pressures were set up within the soil by applying a 0.2

compressive load to the foundation plate over a short time period

0
with no drainage permitted. The foundation was assumed to be 0 0.1 0.2 0.3 0.4 0.5
subjected to a vertical load, and hence average pressure ⌬q 共total x/D

load divided by the cross sectional bearing area of the founda-

tion兲. A nominal foundation load corresponding to an average Fig. 5. Initial excess pore pressure distribution beneath foundation
pressure ⌬q / E⬘ = 0.04 was applied to ensure small strain condi- base plate

24 / INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010

Int. J. Geomech. 2010.10:22-29.

 ing application of a uniform foundation pressure ⌬q for the Smooth interface

D/2
smooth and rough foundations for each of the embedment ratios 1.1
z
considered. 0.4

z/D ~ 1
The distribution of contact pressure beneath the surface foun- 0.3
0.2
dation is largely independent of the interface roughness and the
finite-element prediction of the distribution of contact pressure 0.1

beneath the surface foundations agrees well with the theoretical

stress distribution under a rigid die, given by 共Muki 1961兲
Rough interface

⌬ui 1

冋 册
= 共2兲
⌬q x2
2 1−
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

共D/2兲2

Skirted foundations carry foundation load by bearing beneath the

base plate and skirt tips and friction along the skirts if the inter-
face is rough. For a smooth skirt-soil interface, the entire applied
Surface d/D = 0 d/D = 0.15 d/D = 0.3 d/D = 0.5 d/D = 1
foundation pressure is transmitted to the level of the skirt tips
under one-dimensional conditions resulting in increasing contact Fig. 6. Contours of initial excess pore pressure ratio at intervals of
pressure with increasing embedment ratio. A limiting contact 0.1⌬ui / ⌬q
pressure across the base plate of around 98% of the applied foun-
dation pressure is observed with the remaining load carried by the
skirt tips. As skirt thickness tends to zero the relative contact
pressure across the base will tend to unity. In this study, the skirt kE⬘ 共1 − ␯⬘兲
thickness of t / D = 0.002 represents less than 1% of the cross- cv = 共3兲
␥w 共1 − 2␯⬘兲共1 + ␯⬘兲
sectional area of the foundation such that a relatively small com-
ponent of the applied foundation load is carried by the skirt tips. where k = soil permeability; E⬘ = Young’ s modulus of the soil;
For a rough skirt-soil interface, a portion of the applied founda- ␥w = unit weight of water; and ␯⬘ = Poisson’ s ratio.
tion load is carried by friction along the skirts. The proportion of The first data point of each series in Fig. 7 corresponds to the
the applied load carried by skirt friction increases with increasing end of the initial time increment t = h2␥w / 6Ek 关Eq. 共1兲兴 and the
embedment ratio leading to a reduction in contact pressure with magnitude of the initial excess pore pressures in Fig. 7共a兲, corre-
increasing embedment ratio. The variation in contact pressure sponds to those indicated in Fig. 5 at x / D = 0. The increasing
across the base plate diminishes with increasing embedment ratio, contact pressure with increasing embedment ratio of the smooth
irrespective of interface roughness, as the confinement provided foundations and the decreasing contact pressure with increasing
by the skirts prevents lateral spreading of the applied load. embedment ratio of the rough foundations is again evident.
Fig. 6 shows contours of initial excess pore pressure ratio at The excess pore pressure time histories of the surface founda-
equal intervals of 0.1⌬ui / ⌬q, for each embedment ratio for tions and the rough skirted foundations with embedment ratio up
smooth and rough interface conditions. For the surface founda- to d / D = 0.3 exhibit the characteristic Mandel-Cryer effect 共Man-
tions, a change in initial pore pressure ⌬ui of 10% of the applied del 1950; Cryer 1963兲, with the excess pore pressure increasing
foundation pressure ⌬q is observed at a depth z ⬃ D, consistent above the initial value during the early stage of consolidation. The
with analytical solutions 共Poulos and Davis 1974兲. The extent of Mandel-Cryer effect is a stress transfer effect significant in three-
significant excess pore pressures beneath ground level increases dimensional consolidation. As soil near to the free surface drains
with increasing embedment ratio for both the smooth and rough more quickly than deeper soil, the strain associated with drainage
foundations, as would be expected. The distribution of excess of the nearer surface material “squeezes” the deeper undrained
pore pressure below skirt tip level of the smooth foundations is material leading to an increase in total stress over the initial in-
essentially independent of skirt length. With rough skirts, the ex- crease in total stress due to the externally applied load. Therefore
tent of significant stress changes beneath the skirt tips consistently in the early stages of consolidation all the soil experiences an
reduces with increasing embedment ratio due to the increasing increase in excess pore pressure over the initial value. As the
proportion of the applied load carried by the skirts. drainage front advances the deeper soil drains and total stress
reduces. The early rate of development of effective stress is small,
gradually increasing with time. At the point when the rate of
Excess Pore Pressure Dissipation
development of effective stress becomes greater than the rate of
Figs. 7共a and b兲 show the progress of dissipation of excess pore increase of total stress the excess pore pressure, having achieved
pressures measured along the interface with the base plate at the a maximum, will start to dissipate.
centerline of the foundation. Fig. 7共a兲 shows excess pore pressure The magnitude of the effect and the duration over which the
normalized by the applied foundation pressure ⌬u / ⌬q, and Fig. effect prevails decreases with increasing embedment ratio. Excess
7共b兲 shows excess pore pressure as a proportion of the initial pore pressure behavior during three-dimensional consolidation is
change in pore pressure following application of the applied foun- affected by the magnitude of Poisson’s ratio. The magnitude of
dation pressure ⌬u / ⌬ui. Time is expressed in terms of the dimen- the Mandel-Cryer effect reduces with increasing Poisson’s ratio
sionless time factor T = cvt / D2, where t is the time following 共Schiffmann et al. 1969兲 and as such the embedment ratio at
application of the applied foundation pressure, ⌬q, D is the diam- which the effect would cease to manifest itself is also likely to
eter of the foundation, and cv is the coefficient of consolidation depend on Poisson’s ratio. However, over a practical range of
共Davis and Poulos 1968兲 drained Poisson’s ratio the effect is likely to be minimal.

INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010 / 25

Int. J. Geomech. 2010.10:22-29.

 (a) (a)
1 1
D/2 Smooth d/D D/2
0
z=0
z increasing 0.5
z
0.8 d 0.8 d

z = 0.5D

Smooth d/D Rough 0.6

0.6

∆u/∆q
∆u/∆q

0 z=D

0.15
0.4 0.3 0.4

0.5
1
0.2 0.2
z = 2D

0 0
0.0001 0.001 0.01 0.1 1 10 100
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

0.0001 0.001 0.01 0.1 1 10 100

T = cvt/D2 T=cvt/D2
(b)
(b)
1.2 1.2

1 1

0.8 0.8
Slower dissipation of excess
∆u/∆ui

∆u/∆ui
pore pressure under base
0.6 0.6 plate than at skirt tip level

0.4 0.4

0.2 0.2

0 0
0.0001 0.001 0.01 0.1 1 10 100 0.0001 0.001 0.01 0.1 1 10 100
2
T = cvt/D T = cvt/D2

Fig. 7. Time histories of dissipation of excess pore pressure mea- Fig. 8. Time histories of dissipation of excess pore pressure mea-
sured at the center line of the foundation along the interface with the sured along the center line of the foundation at various depths for
base plate in terms of 共a兲 excess pore pressure ratio ⌬u / ⌬q; 共b兲 smooth foundations d / D = 0 and 0.5 in terms of 共a兲 excess pore pres-
normalized excess pore pressure ⌬u / ⌬ui sure ratio ⌬u / ⌬q; 共b兲 normalized excess pore pressure ⌬u / ⌬ui

For the rough foundations, the time at which dissipation of seen in Fig. 6兲. The time for dissipation of excess pore pressures
excess pore pressure starts and the rate of dissipation is dependent increases with increasing depth beneath the surface foundation
on embedment ratio. Due to the Mandel-Cryer effect, the degree while, for the skirted foundation, slower dissipation of excess
of dissipation of excess pore pressure during the early stage of pore pressures is observed under the baseplate than at skirt tip
consolidation is greater for larger embedment ratios. With time level. The Mandel-Cryer effect is evident at all depths beneath the
the trend reverses and dissipation of excess pore pressures in- smooth surface footing, and as would be expected, the relative
creasingly slows with increasing embedment ratio. The smooth magnitude of this effect increases with increasing depth, and the
skirted foundations, unaffected by the Mandel-Cryer effect, ex- peak advances in time with increasing depth 关most clearly seen in
hibit a consistent increase in the time for dissipation of excess Fig. 8共b兲兴.
pore pressures with increasing embedment ratio. The time for
dissipation of excess pore pressures is increasingly prolonged for
Time-Settlement Response
embedment ratios d / D ⬎ 0.5 as drainage in the soil plug domi-
nates over drainage in the far field. This may be seen from the Fig. 9 shows the immediate settlement wi, final consolidation
response for d / D = 1 共for the smooth case兲, where the time for settlement wcf , and total settlement wt, where total settlement is
50% pore pressure dissipation ⌬u / ⌬ui = 0.5 is T ⬃ 0.43, compared taken as the sum of the immediate and primary consolidation
with the one-dimensional consolidation solution within the soil settlement wt = wi + wcf as a function of embedment ratio d / D. Fig.
plug, where ⌬u / ⌬ui = 0.5 occurs at T ⬃ 0.38. The apparently faster 9共a兲 shows settlement normalized by the foundation diameter, soil
decay of pore pressure for the skirted foundation case 共particu- Young’s modulus, and applied foundation pressure wi, cE / D⌬q,
larly at low values of T兲 is due to additional transfer of load to the and Fig. 9共b兲 shows settlement normalized by final total settle-
caisson tip as consolidation progresses. ment 共wi, cf / wt兲.
Fig. 8 shows the variation of dissipation of excess pore pres- Fig. 9共a兲 shows that the immediate displacement of the smooth
sure as a function of depth below ground level for a selected foundations are virtually independent of embedment ratio, while
example of a smooth skirted foundation with an embedment ratio the displacement of the rough foundations reduces with increas-
d / D = 0.5 compared with the smooth surface case d / D = 0. Fig. ing embedment ratio due to the portion of applied foundation load
8共a兲 shows excess pore pressure normalized by the applied foun- being carried by friction along the skirts. Consolidation settlement
dation pressure, and Fig. 8共b兲 shows excess pore pressure as a of the smooth foundations increases with increasing embedment
proportion of the initial change in pore pressure following appli- ratio consistent with the increasing one-dimensional compression
cation of the applied foundation pressure. The reduction in initial within the skirts. Consolidation settlement of the rough founda-
excess pore pressure with depth is evident from Fig. 8共a兲 共as also tions reduces with increasing embedment ratio consistent with an

26 / INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010

Int. J. Geomech. 2010.10:22-29.

 (a) (a)
T = cvt/D2
d/D 0.0001 0.001 0.01 0.1 1 10 100
0 0.2 0.4 0.6 0.8 1 0
0
0.1
wcf D/2
0.2
0.2
d

0.4 0.3

wcE/D∆q
wi
wE/D∆q

Smooth d/D Rough

0.6 0.4
0
0.15
D/2 0.5
0.8 wt 0.3

0.5
0.6
1
1 d
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

0.7

1.2 (b)
T = cvt/D
2

0.0001 0.001 0.01 0.1 1 10 100

(b) 0
d/D
0 0.2 0.4 0.6 0.8 1
0 0.2

Booker & Small (1986)

0.4
0.2

wc/wcf
wcf 0.6
w/wt

0.4
0.8

0.6
1
wi
Fig. 10. Time histories of consolidation settlement, measured at the
0.8 center line of the foundation along the interface with the base plate in
Smooth Rough
terms of 共a兲 dimensionless settlement ratio wcE / D⌬q; 共b兲 normalized
Fig. 9. Variation of immediate, consolidation, and total settlement settlement wc / wcf
with embedment ratio and interface roughness in terms of 共a兲 dimen-
sionless settlement ratio wE / D⌬q; 共b兲 normalized settlement w / wt
as would be expected. With time, the rate of consolidation quick-
ens for the longer smooth skirted foundations as the consolidation
increasing proportion of the applied load being carried by skirt process is increasingly dominated by pore pressure dissipation in
friction. In line with these trends, total settlement wt共=wi + wcf 兲 of the soil plug and the time histories become more closely banded.
the smooth foundations increases with increasing embedment Comparison of the finite-element results for the surface founda-
ratio while total settlement of the rough foundations reduces with tions with the established analytical solution for a smooth, rigid,
increasing embedment ratio. It is worth noting that in reality soil impermeable surface foundation 共Booker and Small 1986兲 shows
stiffness would usually increase with depth leading to reduced good agreement 关Fig. 10共b兲兴.
settlements, even for smooth sided foundations, as load is trans- Fig. 11 shows the compression in the soil plug measured along
ferred to deeper, stiffer soil. the centerline of the foundation, given by the difference between
Fig. 9共b兲 shows that a greater proportion of total settlement of the final consolidation settlement wcf measured on the interface
the rough skirted foundations is contributed by the immediate with the base plate and at skirt tip level. Compression in the soil
component of settlement rather than the consolidation settlement.
For smooth foundations with low embedment ratios, immediate
settlement is also the major component of the total settlement, but d/D
0 0.2 0.4 0.6 0.8 1
with increasing embedment ratio 共and increasing one-dimensional 0
compression in the soil plug兲 the consolidation settlements make
up an increasing proportion of the total settlement. 0.1
Fig. 10 shows time histories of consolidation settlement wc D/2
0.2
= wt − wi. Fig. 10共a兲 shows settlement normalized by the founda- Compression
wcfE/D∆q

tion diameter, soil Young’s modulus and applied foundation pres- 0.3 in soil plug
d
sure wcE / D⌬q, and Fig. 10共b兲 shows settlement normalized by
0.4
final consolidation settlement wc / wcf . In both cases settlement
was measured on the centerline of the foundation along the inter- 0.5 wcf at
face with the base plate. Fig. 10共a兲 shows the increase in consoli- skirt tip level
0.6
dation settlement with increasing embedment ratio for the smooth wcf under base plate
foundations and reduction in consolidation settlement with in- 0.7
creasing embedment ratio for the rough foundations throughout Smooth Rough
the consolidation process. Fig. 10共b兲 shows a general trend of
increasing consolidation times with increasing embedment ratio, Fig. 11. Compression in soil plug

INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010 / 27

Int. J. Geomech. 2010.10:22-29.

 (a) (a)
1 T = cvt/D2
0.0001 0.001 0.01 0.1 1 10 100
D/2
0.9 0

0.8
d
0.1 D/2
0.7

0.6 Smooth d/D kh/kv 0.2 d

∆u/∆q

0 1
0.5
0.3

wcE/D∆q
1
3 Smooth d/D
0.4
10 0
0.3 0.4 1

0.2 kh/kv
0.5
1
0.1
0.6 3
0
10
0.0001 0.001 0.01 0.1 1 10 100
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

T = cvt/D2 0.7
(b)
1.2 (b)
T = cvt/D2
0.0001 0.001 0.01 0.1 1 10 100
1 0

0.8 0.2
kh/kv increasing
∆u/∆ui

kh/kv increasing
0.6
0.4

wc/wcf
0.4
0.6
0.2

0.8
0
0.0001 0.001 0.01 0.1 1 10 100
T = cvt/D2
1

Fig. 12. Time histories of dissipation of excess pore pressure mea- Fig. 13. Time histories of consolidation settlement, measured at the
sured along the center line of the foundation along the interface with center line of the foundation along the interface with the base plate
the base plate for smooth foundations d / D = 0 and 1 and permeability for smooth foundations d / D = 0 and 1 and permeability ratios kv / kh
ratios kv / kh = 1, 3, and 10 in terms of 共a兲 excess pore pressure ratio = 1, 3, and 10 in terms of 共a兲 dimensionless settlement ratio
⌬u / ⌬q; 共b兲 normalized excess pore pressure ⌬u / ⌬ui wcE / D⌬q; 共b兲 normalized settlement wc / wcf

plug increases with increasing embedment ratio as expected, and

is significantly greater for the smooth foundations; this is consis- with increasing permeability ratio kh / kv as would be expected. It
tent with a greater proportion of the load being carried by the soil is also interesting to note the lessening of the Mandel-Cryer effect
plug, in contrast to rough foundations where an increasing pro- for the surface foundation with increasing permeability aniso-
portion of the foundation load is carried by skirt friction. tropy. Up to a fivefold increase in consolidation time is observed
for the case with permeability ratio kh / kv = 10 compared with the
isotropic case kh / kv = 1 and the effect of embedment ratio seems
Effect of Anisotropic Permeability secondary to the variation in permeability.
The rate of consolidation and variation of excess pore pressures
with time depend on the flow properties of the soil, which will
generally be anisotropic. Booker and Randolph 共1984兲 presented Concluding Remarks
an analytical solution for uniform circular and rectangular loading
of an infinitely deep layer with cross-anisotropic permeability and This study considered the idealized elastic consolidation response
indicated a significant effect on the rate of consolidation of the beneath circular skirted foundations as a function of embedment
ratio of horizontal to vertical permeability of the soil. Selected 共or skirt length兲 ratio and skirt-soil interface roughness. The study
cases of the smooth surface foundation and the smooth skirted has shown that the stress distribution around a skirted foundation
foundation with embedment ratio d / D = 1 have been considered and within the soil plug following application of a surface load is
with permeability ratios kh / kv of 3 and 10. Results from the an- highly dependent on skirt-soil interface roughness. Smooth skirts
isotropic analyses are compared with the results from the isotro- result in the surface load being transmitted to the level of the skirt
pic case kh / kv = 1 in Figs. 12 and 13. tips under one-dimensional conditions resulting in increasing con-
Fig. 12 shows time histories of excess pore pressure dissipa- tact pressure with increasing embedment ratio. Conversely, with
tion, normalized by the applied foundation pressure ⌬u / ⌬q and rough skirts the proportion of the applied load carried by skirt
as a proportion of the initial value ⌬u / ⌬ui. Fig. 13 shows time friction increases with increasing embedment ratio leading to a
histories of settlement, normalized by foundation diameter, reduction in contact pressure with increasing embedment ratio.
Young’s modulus, and applied foundation pressure 共wcE / D⌬q兲, Consolidation times are increasingly prolonged with increasing
and as a proportion of the final consolidation settlement 共wc / wcf 兲. embedment ratios and for both smooth and rough skirt-soil inter-
The magnitude of initial excess pore pressure and final settle- face conditions, additional load is transferred to the skirt tip, and
ment is independent of the permeability characteristics, while the in the case of the rough skirts to the shaft, as consolidation
rate of excess pore pressure dissipation and settlement increases progresses.

28 / INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010

Int. J. Geomech. 2010.10:22-29.

 The study presented in this paper was intended to assess the uses sous une charge.” Annaies de la Societe Scientific de Bruxelles,
effect of embedment ratio and skirt-soil interface roughness on Series B, 55, 110–113.
the mechanisms of consolidation beneath skirted foundations sub- Biot, M. A. 共1956兲. “General solutions of the equations of elasticity and
jected to uniaxial compressive load. A set of benchmark solutions consolidation for a porous material.” ASME Trans. J. Appl. Mech.,
78, 91–96.
for idealized conditions have been presented, offering an insight
Booker, J. R. 共1974兲. “The consolidation of a finite layer subject to sur-
into the principal mechanisms governing the consolidation re-
face loading.” Int. J. Solids Struct., 10, 1053–1065.
sponse and providing a starting point for further investigation Booker, J. R., and Randolph, M. F. 共1984兲. “Consolidation of a cross-
representing more realistic soil conditions. anisotropic soil medium.” Q. J. Mech. Appl. Math., 37共3兲, 479–495.
Booker, J. R., and Small, J. C. 共1986兲. “The behaviour of an impermeable
flexible raft on a deep layer of consolidating soil.” Int. J. Numer.
Acknowledgments Analyt. Meth. Geomech., 10共3兲, 311–327.
Bye, A., Erbrich, C., Rognlien, B., and Tjelta, T. I. 共1995兲 “Geotechnical
The work described here forms part of the activities of the Centre design of bucket foundations.” Proc., Annual Offshore Technology
Downloaded from ascelibrary.org by ENPC: Couperin Consortium on 09/17/14. Copyright ASCE. For personal use only; all rights reserved.

for Offshore Foundation Systems, established under the Austra- Conf., OTC, Houston.
lian Research Council’s Research Centres Program, and now sup- Chiarella, C., and Booker, J. R. 共1975兲. “The time-settlement behaviour
ported through Grant Nos. FF0561473 and DP0988904 and of a rigid die resting on a deep clay layer.” Int. J. Numer. Analyt.
Centre of Excellence funding from the State of Western Australia. Meth. Geomech., 8, 343–357.
This support is gratefully acknowledged. Clukey, E. C., Templeton, J. S., Randolph, M. F., and Phillips, R. A.
共2004兲. “Suction caisson response under sustained loop-current
loads.” Proc., Offshore Tech. Conf., OTC, Houston.
Cryer, C. W. 共1963兲. “A comparison of the three dimensional consolida-
Notation tion theories of Biot and Terzaghi.” Q. J. Mech. Appl. Math., 16,共4兲
401–412.
The following symbols are used in this paper: Davis, E. H., and Poulos, H. G. 共1968兲. “The use of elastic theory for
A ⫽ cross-sectional 共bearing兲 area of foundation settlement prediction under three-dimensional conditions.” Geotech-
共=␲D2 / 4兲; nique, 18共1兲, 67–91.
cv ⫽ coefficient of consolidation 关see Eq. 共3兲兴; Gibson, R. R., Schiffman, R. L., and Pu, S. L. 共1970兲. “Plane strain and
D ⫽ foundation diameter; axially symmetric consolidation of a clay layer on a smooth impervi-
d ⫽ skirt embedment depth; ous base.” Q. J. Mech. Appl. Math., 23共4兲, 505–519.
d / D ⫽ embedment ratio; Hansen, B., Nowacki, F., Skomedal, E., and Hermstad, J. 共1992兲 “Foun-
E⬘ ⫽ drained Young’s modulus; dation design: Troll platform.” Proc., Int. Conf. on Behaviour of Off-
kh and kv ⫽ coefficient of horizontal and vertical shore Structures, BOSS ‘92, Bentham, London, 921–936.
permeability; Humpheson, C. 共1998兲. “Foundation design of Wandoo B concrete grav-
ity structure: Offshore site investigation and foundation behaviour.”
T ⫽ dimensionless time factor 共=cvt / D2兲;
Soc. Underwat. Technol., 1, 353–367.
t ⫽ skirt thickness; Mandel, J. 共1950兲. “Etude mathiematique de la consolidation des sols.”
t / D ⫽ skirt thickness ratio; Actes du colloque international de mechanique, Vol. 4, Poitier,
V ⫽ applied foundation load; France, 9–19.
wc ⫽ consolidation settlement; McNamee, J., and Gibson, R. E. 共1960兲. “Plane strain and axially sym-
wc / wcf ⫽ normalized consolidation settlement; metric problems of the consolidation of a semi-infinite clay stratum.”
wcf ⫽ final consolidation settlement; Q. J. Mech. Appl. Math., 13共2兲, 210–227.
wE / D⌬q ⫽ dimensionless settlement ratio; Muki, R. 共1961兲. “Axisymmetric problems of the theory of elasticity for
w f ⫽ total final settlement 共=wi + wc兲; a semi-infinite solid and thick plate.” Progress in solid mechanics,
wi ⫽ immediate settlement; Vol. 1, North-Holland, Amsterdam.
wi / w f ⫽ normalized immediate settlement; Poulos, H. G., and Davis, E. H. 共1974兲. Elastic solutions for soil and rock
⌬q ⫽ applied foundation pressure 共=V / A兲; mechanics, Wiley, New York.
⌬u ⫽ excess pore-water pressure; Schiffmann, R. L., Chen, A. T. F., and Jordan, J. 共1969兲. “An analysis of
consolidation theories.” J. Soil Mech. Found. Div., 95共SM1兲, 285–
⌬ui ⫽ initial excess pore-water pressure;
313.
⌬u / ⌬q ⫽ dimensionless excess pore pressure ratio; Støve, O. J., Bysveen, S., and Christophersen, H. P. 共1992兲. “New foun-
⌬u / ⌬ui ⫽ normalized excess pore pressure; dation systems for the Snorre development.” Proc., Annual Offshore
⌬␴⬘v / ⌬ui ⫽ dimensionless effective stress ratio; Technology Conf., OTC, Houston.
⌬␴⬘v / ⌬q ⫽ dimensionless effective stress ratio; Tjelta, T. I., Aas, P. M., Hermstad, J., and Andenaes, E. 共1990兲. “The skirt
␥w ⫽ unit weight of water; and piled Gullfaks C platform installation.” Proc., Annual Offshore Tech-
␯⬘ ⫽ drained Poisson’s ratio. nology Conf., OTC, Houston.
Vermeer, P. A., and Verruijt, A. 共1981兲. “An accuracy condition for con-
solidation by finite elements.” Int. J. Numer. Analyt. Meth. Geomech.,
References 1, 1–14.
Watson, P. G., and Humpheson, C. 共2007兲. “Foundation design and in-
ABAQUS analysis users’ manual; version 6.8. 共2008兲. Dassualt Systèmes stallation of the Yolla A platform.” Proc., 6th Int. Offshore Site Inves-
Simula Corp., Providence, R.I. tigation and Geotechnics Conf. SUT, Society for Underwater
Biot, M. A. 共1935兲. “Le problem de la consolidation des matieres argile- Technology, London, 399–412.

INTERNATIONAL JOURNAL OF GEOMECHANICS © ASCE / JANUARY/FEBRUARY 2010 / 29

Int. J. Geomech. 2010.10:22-29.