You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/260417358

Modified expression for the failure criterion of Bucket foundations subjected to


combined loading

Article  in  Canadian Geotechnical Journal · February 2014


DOI: 10.1139/cgj-2012-0308

CITATIONS READS
35 180

3 authors, including:

Lars Bo Ibsen Amin Barari


Aalborg University RMIT University
178 PUBLICATIONS   2,348 CITATIONS    151 PUBLICATIONS   2,546 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Analysis of liquefaction-induced bearing capacity degradation of suction caissons View project

Wave-induced liquefaction around offshore foundations View project

All content following this page was uploaded by Amin Barari on 09 July 2021.

The user has requested enhancement of the downloaded file.


1250

ARTICLE
Modified expression for the failure criterion of bucket
foundations subjected to combined loading
K.A. Larsen, L.B. Ibsen, and A. Barari
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

Abstract: Recently, various loading tests with small- and large-scale bucket foundations were performed on buckets of varying
sizes, embedment ratios, and load paths with saturated dense Aalborg University Sand No. 1 at the geotechnical laboratory of
Aalborg University. In the present study, the capacity and behavior of bucket foundations subjected to combined loads were
investigated on the basis of the results of these experimental studies. Although previous authors had proposed a linear
relationship between the moment capacity and the vertical load, this relationship did not fit the observed capacities in the
present study. Therefore, a modified expression is proposed to describe the combined capacity as a function of the tensile
capacity and the inclination factor.

Key words: bucket foundation, combined loading, failure criterion, tensile vertical capacity.

Résumé : Récemment, différents essais de chargement ont été réalisés sur des fondations avec caissons à succion de petite et
grande taille au laboratoire de géotechnique de l’Université d’Aalborg. Ces essais ont permis de varier la taille des caissons, le
ratio d’enfouissement et les chemins de charge, en utilisant du sable dense et saturé de type no.1 de l’Université d’Aalborg. Dans
cette étude, la capacité et le comportement des fondations avec caisson de succion soumises à des charges combinées ont été
étudiés à partir des résultats des études expérimentales. Des auteurs précédents avaient proposé une relation linéaire entre le
moment de capacité et la charge verticale; cependant cette relation ne correspondait pas avec les capacités observées dans cette
For personal use only.

présente étude. Alors, une expression modifiée est proposée afin de décrire la capacité combinée en fonction de la capacité en
tension et du facteur d’inclinaison. [Traduit par la Rédaction]

Mots-clés : fondation avec caisson à succion, charge combinée, critère de rupture, capacité verticale en tension.

Introduction tion under combined loading. However, the validity of the super-
Originally developed for the offshore oil and gas industry, position of solutions for inclination and eccentricity to represent
general loading has come under scrutiny in light of, for example,
bucket foundations are currently being used as a new solution for
the bound solutions reported by Ukritchon et al. (1998), and sim-
offshore wind turbines. Recent investigations of the maximum
ilar shortcomings for circular foundations confirmed by Gourvenec
permissible load that can be sustained by offshore foundations for
and Randolph (2003). Other studies have confirmed that by ne-
wind turbines (known as the “bearing capacity”) have attracted
glecting the tensile capacity, classical solutions for offshore de-
much attention (Barari and Ibsen 2012). Traditionally, the bearing
sign provide conservative solutions. In reality, the tensile capacity
capacity has been understood as the “vertical capacity,” and sev-
may be mobilized by some types of offshore shallow foundations,
eral expressions for bearing capacity factors have been proposed
such as offshore bucket foundations for wind turbines.
(Ibsen et al. 2012). Meyerhof (1951, 1953), Hansen (1970), and Vesic
Suction caissons are large cylindrical foundation units installed
(1975) initially examined the behavior of foundations under con- by active suction after partial penetration by deadweight. These
ditions of both horizontal and vertical loading as well as large units have been extensively investigated by numerous authors
overturning moments. (Clukey et al. 1995; El-Gharbawy et al. 1997; Erbrich and Tjelta
Alternatives to piled foundations, such as shallow foundations, 1999; El-Gharbawy and Olson 2000; Iskander et al. 2002; El-Sherbiny
are increasingly being used as viable options for fixed platforms, et al. 2005; Luke et al. 2005; Vásquez et al. 2010). In an early study,
particularly in the development of marginal fields. However, Erbrich and Tjelta (1999) performed an extensive series of model
there are no accepted guidelines for shallow foundations. Design tests and finite element analyses to describe the important
guidelines for estimating the bearing capacity of offshore shallow geotechnical principles related to the installation of bucket foun-
foundations set out by the DNV (1992), API (2000), and ISO (2000) dations in dense sand when suction pressures are applied to in-
have the same roots as those adopted for onshore design, despite crease the static driving force and degrade the penetration
their large differences in loading conditions. Nevertheless, all resistance. Iskander et al. (2002) studied model caissons installed
of these guidelines are ultimately based on classical bearing capacity on sand and clay in the laboratory, and found that suction pres-
theories whose applicability to offshore design is questionable. sure may serve as a viable alternative to traditional techniques for
Traditional theories use modification factors to extend the ba- caisson installation. They measured the small positive pore-water
sic solution for the failure of a vertically loaded strip foundation pressures generated during the installation procedure by suction
resting on the surface of a uniform deposit to the more complex pressure and pushing. Houlsby et al. (2005) showed that if the
(and realistic) condition of a three-dimensional skirted founda- tensile load is applied very slowly, then the pore pressures will be

Received 16 August 2012. Accepted 23 September 2013.


K.A. Larsen. Cowi A/S, Aalborg, Denmark.
L.B. Ibsen and A. Barari. Department of Civil Engineering, Aalborg University, Sohngårdsholmsvej 57, 9000 Aalborg, Aalborg, Denmark.
Corresponding author: A. Barari (e-mail: ab@civil.aau.dk).

Can. Geotech. J. 50: 1250–1259 (2013) dx.doi.org/10.1139/cgj-2012-0308 Published at www.nrcresearchpress.com/cgj on 2 October 2013.
Larsen et al. 1251

small and a fully drained calculation can be used to calculate the Fig. 1. Geometry considered by Houlsby et al. (2005).
capacity. Clukey et al. (1995) performed seven centrifuge tests to
evaluate the behavior of steel suction caissons under uplift load-
ings and soil conditions typical of the deepwater Gulf of Mexico.
They established the relationships between the ratio of the cyclic
load to the static uplift resistance and the number of applied
cycles to cause failure of the foundation. Later, Luke et al. (2005)
systematically studied the behavior of caissons that were installed
in the laboratory in normally consolidated clay by using dead
weight and suction. They obtained ranges for the external side
resistance factor (␣) of 0.5–0.8 and for the reverse end bearing
factor (Nc) of 13–21 for the case of rapid pullout.
Among the offshore shallow foundations, the suction bucket
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

foundation system is unique for its ease of installation and reus- Fig. 2. Illustration of failure surface, according to Butterfield and
ability (Erbrich and Tjelta 1999; Ibsen et al. 2004). A bucket foun- Ticof (1979).
dation is essentially an upside-down bucket-like steel structure
that is lowered to the seabed under its own weight and then
inserted into the soil by pumping the water from the bucket’s
interior. Other advantages of bucket foundations include reduced
construction costs and the availability of passive suction forces to
counteract uplift. Bucket foundations are effective under suitable
soil conditions (e.g., fine sand or clay materials) in water depths
from near-shore to approximately 55 m (Feld 2001; Ibsen et al.
2004; Achmus and Abdel-Rahman 2005; Ibsen 2008).
In 1999, Sukumaran and McCarron (1999) reported the axial and
lateral capacities of suction bucket foundations founded on soft
clays, as determined by numerical analysis. Later research per-
formed at Aalborg University (Barari and Ibsen 2011, 2012, 2013; Fig. 3. Horizontal loading of bucket foundation.
Ibsen and Barari 2011) reported the experimental and numerical
For personal use only.

results of vertical and moment loading on small-scale circular


surface and bucket foundation models installed on Baltic clay. A
series of parametric studies were performed by varying the soil
parameters E and cu in numerical simulations. The tangent elastic
modulus corresponding to the 25% strain E25 achieved the best fit
with the experimental behavior. Wang and Jin (2008) and Bransby
and Yun (2009) used the finite element method to study the effect
of moment on the bearing capacity of bucket foundations. In
addition, Aubeny et al. (2003) estimated the lateral loading capac-
ity of bucket anchors by varying the load reference point and load
inclination angle.
The present research employed a special apparatus developed
at Aalborg University to investigate the behavior of bucket foun-
dations installed on saturated dense Aalborg University Sand No.
1 when exposed to combined loads. Tests were performed by load-
ing the buckets with a load or deformation path that allowed the
bucket to move freely, similar to foundations in natural condi-
tions. Several tests were performed on bucket foundations ex-
posed to low vertical loads. The bucket diameter was typically
300 mm, with varying embedment ratios. Tests performed in the
laboratory involved loads comparable to loading from offshore on bucket foundations performed at the geotechnical laboratory
wind turbines. The applied vertical load was low relative to the of Aalborg University are reported in Ibsen et al. (2012).
vertical bearing capacity of the foundation.
Combined peak capacity
Analysis of the bearing capacity Experiments carried out at Aalborg University are especially
Vertical bearing capacity relevant for considering the drained combined capacity of bucket
A special case of combined loading is the situation in which the foundations in sand. Two failure criteria were considered in the
foundation is loaded purely in the vertical direction (H = M = 0). conducted research. The first criterion, which describes the capac-
The purely vertical bearing capacity, Vpeak, can be calculated from ity at low vertical load, is presented in the current work. The
the general bearing capacity formula proposed by Prandtl (1921) second criterion, which proposes the full capacity for 200 mm
and Terzaghi (1943) in the following form: buckets with an embedment ratio varying from 0 to 1, will be
illustrated according to the macro model approach in subsequent

共21 B␥ N 兲
studies.
(1) Vpeak ⫽ ␥ ⫹ cNc ⫹ qNq A
Tensile vertical capacity
The apex of the failure surface (i.e., the intersection with the
where Vpeak is the peak vertical load, B is the foundation width, V-axis) was taken as the tensile capacity of the bucket foundation.
and A is the foundation area. The results of vertical capacity tests If the hardening of the yield surface is assumed to be isotropic and

Published by NRC Research Press


1252 Can. Geotech. J. Vol. 50, 2013

Fig. 4. Loading setup designed at Aalborg University for bucket foundations.


Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

the lower apex is constant, then the apex of the failure surface can Fig. 5. Spider 8 sampling device connected to a PC.
For personal use only.

be taken as the tensile capacity for an expanding yield surface. It


is reasonable to assume that the vertical, moment, and horizontal
capacities of a bucket foundation are similar to those of an em-
bedded foundation. However, the tensile capacities of these foun-
dations must be different. The resistance on the bucket is
evaluated as the sum of the frictions on the outside and inside of
the skirt. The tensile capacity may be calculated by assuming a
limiting shear stress between the soil and skirt, as described by
the Coulomb failure criterion.
The drained pullout capacity, Q, of a cylindrical caisson may be
expressed as follows (Iskander et al. 2002):

(2) Q ⫽ Wc ⫹ Q s

where Wc is the sum of the submerged weights of the caisson,


templates, and ballast, and Q s is the frictional resistance on both
the inside and outside of the caisson wall. Houlsby et al. (2005) When an upwards-directed bucket is subjected to a vertical
showed that the effective stresses on the annular rim may be force, the stresses near the skirt are reduced due to the frictional
neglected, as they are likely to be small. Therefore, the tensile forces further up the skirt. Equation (4) does not account for this
capacity of bucket foundations provided by foundation skirts, change and, consequently, overestimates the force that can be
accounting for the effects of stress enhancement, can be obtained developed. To account for the reduction in stress, Houlsby et al.
as follows: (2005) suggested that the tensile capacity be calculated with the
following formulae:

(3) Vt ⫽ 冕
0
h
␴v0 dz(K tan␦)o(␲Do) ⫹ 冕
0
h
␴v0 dz(K tan␦)i(␲Di)
Vt ⫽ ⫺ ␥ Zo2 y0 冉Zh 冊(K tan␦) (␲D ) ⫺ ␥ Z y 冉Zh 冊(K tan␦) (␲D )
o o
 2
i i i i
o i
D
where K is the lateral earth pressure coefficient, and ␦ is the fric- Zi ⫽
4(K tan␦)i
tion angle between the skirt and the soil. The z variation, which (5)
Do(m2 ⫺ 1)
ultimately takes a value of h as the penetration depth, is apparent Zo ⫽
from Fig. 1. The indices o and i indicate the outside and inside of 4(K tan␦)o
the bucket, respectively.
If the reduction of the vertical stress close to the bucket is not
y共Zh 兲 ⫽ e ⫺(h/Z)
⫺1⫹ 共Zh 兲
taken into account, then the tensile vertical load on the bucket for
a penetration depth of h is given by: where m is the distance over which the stresses are reduced out-
side the foundation relative to the outer diameter (i.e., mDo; Fig. 1).
␥ h2 ␥ h2 Houlsby et al. (2005) suggested values of m and K tan␦ equal to 1.5
(4) Vt ⫽ ⫺ (K tan␦)o(␲Do) ⫺ (K tan␦)i(␲Di)
2 2 and 0.7, respectively, from an analysis of laboratory tests given in

Published by NRC Research Press


Larsen et al. 1253

Fig. 6. Geometry of the test box designed for small-scale tests.


Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

Fig. 7. Installation of bucket foundation using a hydraulic cylinder.


Kelly et al. (2004) on dense sand (Dr = 80%). Alternatively, Houlsby
et al. (2006) suggested values of m and K tan␦ of 2.0 and 1, respec-
tively, from a large-scale test on dense sand (Dr = 80%–85%).
The tensile capacity presented in eqs. (4) and (5) is only valid when
the contribution from the inside friction does not exceed the weight
of the soil within the bucket (i.e., the soil plug). Therefore, in this
case, the inside friction is replaced by the weight of the soil plug.

Failure surface
For personal use only.

Failure envelopes are particularly relevant to offshore shallow


foundations, which are typically three-dimensional, embedded,
founded in soils, and subject to significant horizontal loading and
moment, owing to harsh environmental forces (Gourvenec and
Barnett 2011). The idea of a three-dimensional failure surface was
originally suggested by Roscoe and Schofield (1956). Butterfield
and Ticof (1979), who were among the first to investigate the ca-
pacity of strip footings on dense sand, found that an elliptical
Fig. 8. Distribution of grains for Aalborg University Sand No. 1.
cigar-shaped failure surface could capture the measured data of a
strip footing (both surface and embedded) (Fig. 2).
Failure envelopes have been derived for various shallow foun-
dations, including strip, rectangular, and circular foundations, as
well as spudcans with zero- and full-tension foundation–soil inter-
faces (Taiebat and Carter 2002; Gourvenec and Randolph 2003;
Gourvenec 2007). In radial planes denoted by (M/D; H) or (M/DVpeak;
H/Vpeak) planes, the failure surface behaves as a rotated ellipsoid
(Butterfield 1981; Georgiadis and Butterfield 1988; Gottardi and
Butterfield 1993; Cassidy 1999; Gottardi et al. 1999). However, few
failure envelopes have been presented by a closed-form solution,
f(V, H, M/D), to make them accessible for routine design use.

Laboratory work
Test procedure
To investigate the drained behavior of bucket foundations on
cohesionless soil, more than 100 small-scale tests on bucket foun-
dations subjected to static combined loads were recently per-
formed with water-saturated dense Aalborg University Sand No. 1
at Aalborg University. Two types of loading tests, including pure
vertical loading (Ibsen et al. 2012) and combined loading, were
performed on bucket foundations of various diameters and skirt Table 1. Values of strength parameters for Aalborg
lengths. Tests with bucket foundations subjected to combined University Sand No. 1.
loads were performed with a constant vertical load (V) and a con-
 
stant ratio of moment to horizontal load (M/H). Different load e ␸t,a (°) ct,a (kPa)
paths (M/H) were applied to the foundation via a rigid loading 0.55 41 19.9
tower bolted to the bucket lid (Fig. 3). The load path given by M/DH 0.61 38.6 34.57
varied between 0.37 and 8.7, corresponding to loads from waves 0.7 34.21 40.27
and wind, respectively. Figure 4 shows the loading setup designed 0.85 30.93 7
at Aalborg University for bucket foundations. The loading velocity

Published by NRC Research Press


1254 Can. Geotech. J. Vol. 50, 2013

Fig. 9. Variation of triaxial friction angle for Aalborg University Sand No. 1. Results from triaxial tests are shown as points.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

was maintained during the tests at a rate that ensured a drained


response.
Displacements and forces during the tests were measured with
(6) q  ⫽ ␴1 ⫺ ␴3 ⫽
sin共␸t,a


1 ⫺ sin共␸t,a

兲 冋
␴3 1 ⫹

ct,a cot共␸t,a
m␴3


册 m

a Spider 8 sampling device connected to a PC (Fig. 5). Loading tests


were carried out in a specially designed rigid steel test box of where ␴1 and ␴3 are the major and minor effective principal stress
1600 mm × 1600 mm × 650 mm (Fig. 6). The sand layer depth was at failure, respectively, q= is the deviatoric stress, and c= is the
increased to approximately 530 mm through the use of a specially effective cohesion. The subscript “t,a” denotes the tangent param-
designed drainage system. eter at high stress levels. The parameter m describes the curvature
Preparation of the test box of the failure envelope at low stress levels. The envelope surface
To ensure uniform load transmission, the ground was made given by eq. (6) was fitted to the results from the triaxial tests. The
For personal use only.

level before each procedure. A wooden frame was placed on the triaxial secant friction angle for Aalborg University Sand No. 1
test configuration to hold the additional sand and water during may be determined from Fig. 9 if the stress level of the problem is
the preparation procedure. The soil surface was raised by applying known. The value of the friction angle for intermediate values of
an extra amount of sand to the test box. Homogeneity was en- the void ratio may be estimated by interpolation. The dilation
sured by increasing the water level while loosening the sand. The angle determined from the triaxial tests at failure is shown in
water flow was raised 5–10 cm above the initial level of the sand Fig. 10. The dilation angle varies with both the void ratio and the
through perforated pipes in the drainage layer, by using a pres- stress level.
sure level just under the critical gradient. The sand was systemat-
ically vibrated twice with a rod vibrator. This vibration procedure Linear failure criteria at low vertical load
was repeated, which lowered the water level and allowed the From a limited number of experiments on bucket foundations
surface to be adjusted to the final level, after the sand was again located in loose dry sand (Dr = 20%), Byrne et al. (2003) found an
loosened by raising the water level through the drainage system. approximately linear relationship between the applied loads at
Excess water was removed through a hole in the wooden frame, to failure. Their experiments were performed with a small V/Vpeak
reduce the water level to the top of the test box. The water was ratio and a single bucket with a diameter of 293 mm and skirt
finally lowered beneath the final surface of the sand through the length of 150 mm. The M/(DH) values during the tests were held
drainage layer with a minimum pressure level. constant at 0.5, 1, and 2, corresponding to impact heights of 150,
The drainage system included a set of perforated pipes, to lead 300, and 600 mm, respectively. These authors did not specify the
water into and out of the test box, stones (2–5 mm in diameter) density or type of sand that they used; therefore, the exact value of
surrounding the drainage pipes, and a sheet of geotextile, placed V/Vpeak is unknown.
between the bottom of the sand layer and top of the drainage They concluded that the following relationship would fit their
layer, to prevent sand from penetrating to the drainage layer. experimental observations:
Installation of the bucket foundation
M ␲D2
For the case in which the skirt length was zero (i.e., a circular (7) f1 ⫹ f2H ⫽ f3␥ 
surface foundation), the foundation was manually located on the D 4
soil surface prior to loading. The installation procedure for bucket
foundations may be performed by applying pressure on the top The equation above was simplified based on a limited number
plate with either a hydraulic cylinder or electric motor (Fig. 7). of tests with d/D = 0.5:

Aalborg University Sand No.1


The properties of Aalborg University Sand No. 1 are well known. (8)
M
D
⫽ f1 ⫹ 冉
f2
k
冊 ⫺1
(V ⫹ f3W)
The grading curve is shown in Fig. 8. The grain density and the
maximum and minimum void ratios are Gs = 2.64, emax = 0.858,
and emin = 0.549, respectively. Triaxial tests with different densi- where k = M/DH, and W ⫽ ␥ 共␲D2/4兲 represents the weight of the
ties of Aalborg University Sand No. 1 have been performed, pro- soil plug inside the bucket. The constants fi in eq. (8) were found to
viding information on mechanical properties. The parameters are be independent of k, although this result was supported by a
listed in Table 1 for void ratios of 0.55, 0.61, 0.7, and 0.85. limited number of experiments.
The relationship proposed by Jacobsen (1970) was used to find Equation (8) was shown to yield linear failure criteria in the
the strength parameters of the sand: planes along the V-axis and in the radial planes for constant values

Published by NRC Research Press


Larsen et al. 1255

Fig. 10. Dilation angle determined from triaxial tests on Aalborg Fig. 11. Failure values from tests with low V/Vpeak values and a
University Sand No. 1. height of impact of (a) h = 2610 mm, (b) h = 1740 mm, and
(c) h = 110 mm.

(a) 1400
d/D=0.25
1200 d/D=0.5
d/D=0.75
d/D=1
1000
Byrne et al. (2003) d/D=0.5

800

M/D [N]
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

600

400
of fi. The parameters that fit the observations were f1 = 3.03, f2 = 1,
and f3 = 0.64, where f3 represented the proportion of the soil plug
200
weight mobilized under the action of moment loading.
Based on the results from Byrne (2000) and Byrne et al. (2003) on
bucket foundations, Byrne and Houlsby (2003) suggested f1 = 3.26, 0
0 200 400 600 800 1000
f2 = 1.073, and f3 = 0.71. However, because the tests were performed V [N]
on loose and dense sand, respectively, the failure parameters are
not expected to be identical in the two cases, due to the depen-
dency of Dr. From these observations, it is clear that any vertical (b) 1200
ballast on the structure has a great impact on the moment bearing
capacity, especially at low values of V/Vpeak (Byrne et al. 2003). d/D=0.25
d/D=0.5
1000 d/D=0.75
Small-scale test results
For personal use only.

d/D=1
In the present study, the relation between the moment capacity Byrne et al. (2003) d/D=0.5
and the applied vertical load is investigated by using the results 800

from laboratory tests on bucket foundations carried out at Aal-


M/D [N]

borg University (Fig. 11). Figure 12 shows the capacities in the


600
radial plane, compared with the relationship in eq. (8). The param-
eters used are those given by Byrne and Houlsby (2003) for bucket
foundations with an embedment ratio of 0.5. 400
The figures show that eq. (8) does not fit the observed capacities
with the given parameters. The failure surface intersects the
V-axis at a value corresponding to f3W, which physically must 200

correspond to the vertical tension capacity.


To describe the combined capacity as a function of the tensile
0
capacity, it is recommended that Vt be used rather than f3W. The 0 200 400 600 800 1000
following relationship is derived to capture the capacities ob- V [N]
served in the laboratory:

M (c)
⫽ ␣(V ⫺ Vt) 450
D d/D=0.25
(9)

␣ ⫽ f1(k) ⫹ f2(k)
d
D 兴 400 d/D=0.5
d/D=0.75
d/D=1
350
Byrne et al. (2003) d/D=0.5
where ␣ is the inclination factor and Vt is the vertical tension 300
capacity, according to the standardized sign convention adopted.
Equation (9) is calibrated to the experimental data and shown 250
M/D [N]

along the V-axis in Fig. 13. The tensile capacity was also evaluated
from eq. (4). Alternatively, the tensile capacity may be calculated 200
by using the approach proposed by Houlsby et al. (2005) and sum-
150
marized in eq. (5), in which the tension takes into account the
reduction in stresses close to the skirt. However, the first method 100
fits the measured capacities at a much higher accuracy, with a
value of K tan(␦) = 2. 50
The relationship proposed in eq. (9) suitably captures the ob-
served capacities. The inclination factor ␣ is observed to be a 0
0 200 400 600 800 1000
function of both the embedment ratio and the impact height k. V [N]
Figure 14 shows the values of ␣ that fit the experiments. A linear
relationship is found between ␣ and the embedment ratio, as is
assumed in eq. (9). However, the relationship between ␣ and the
impact height k is less apparent from the figure: as k increases, ␣

Published by NRC Research Press


1256 Can. Geotech. J. Vol. 50, 2013

Fig. 12. Failure values from tests with low V/Vpeak values and a Fig. 13. Proposed failure criterion compared to experimental failure
vertical constant load of (a) V = 184 N and (b) V = 1000 N. values for tests with low V/Vpeak values and impact heights of
(a) h = 2610 mm, (b) h = 1740 mm, and (c) h = 110 mm.
(a) 900 (a) 1400
d/D=0.25 d/D=0.25
800
d/D=0.5 d/D=0.5
1200
d/D=0.75
700 d/D=0.75
d/D=1
d/D=1
1000
600 Byrne et al. (2003) d/D=0.5

500 800
M/D [N]

M/D [N]
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

400 600

300
400
200
200
100

0
0 -1500 -1000 -500 0 500 1000
0 200 400 600 800 1000
V [N]
H [N]

(b) (b) 1200


1400
d/D=0.25
d/D=0.5
1000
1200 d/D=0.75
For personal use only.

d/D=0.25 d/D=1
d/D=0.5
800
1000 d/D=0.75
d/D=1
M/D [N]

Byrne et al. (2003) d/D=0.5 600


800
M/D [N]

600 400

400 200

200
0
-1500 -1000 -500 0 500 1000
V [N]
0
0 200 400 600 800 1000 1200 1400
H [N]
(c) 500
d/D=0.25
appears to increase towards an asymptotic value. Additional ex- 450 d/D=0.5
periments are needed to examine this variation further. d/D=0.75
Table 2 shows the parameters f1 and f2, determined from the 400
d/D=1
inclination factors ␣ in the proposed failure criterion. The pa-
350
rameters are also illustrated in Fig. 15 with a proposed interme-
diate variation. The parameters tend towards an asymptotic 300
value at high values of k. This finding likely arises from the
M/D [N]

change of the failure mechanism that, for a given impact height, 250

is dominated by the moment and becomes constant. The failure


200
parameters in eq. (9) are functions of k. Thus, the relationship
in the radial plane is nonlinear. Figure 16 shows the modified 150
failure criteria in the radial plane with the variation of the
failure parameters proposed in Fig. 15. The proposed failure 100

criterion captures the measured failure values well and pro- 50


vides a reliable intermediate variation. However, at low H val-
ues, the fitted expression in Fig. 16 exhibits more curvature 0
-1500 -1000 -500 0 500 1000
than the expected results from the literature. V [N]
The proposed failure criterion in eq. (9) suitably captures the
combined capacity of bucket foundations in saturated dense sand
exposed to a small vertical load. Few results from similar tests are
available in the literature. Byrne et al. (2003) performed tests on

Published by NRC Research Press


Larsen et al. 1257

Fig. 14. Variation of measured failure parameter ␣. Fig. 16. Measured loads at failure in radial plane, for tests with
vertical loads of (a) 184 N and (b) 1000 N. Proposed failure criterion is
shown as dotted lines.

(a) 900
d/D=0.25
800 d/D=0.5
d/D=0.75
700 d/D=1
New Failure Criteria
600
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

500

M/D [N]
400

300

200

100

0
0 200 400 600 800 1000
H [N]

Table 2. Fitted failure parameters. (b) 1400


h (mm) k = h/D f1 f2
For personal use only.

d/D=0.25
110 0.37 0.167 0.053 d/D=0.5
1200
1740 5.8 0.283 0.267 d/D=0.75
2610 8.7 0.283 0.267 d/D=1
1000 New Failure Criteria

Fig. 15. Parameters for proposed failure criterion given in Table 2.


Possible variations of f1 and f2 are also shown. 800
M/D [N]

0.3
600
0.25

0.2 400
f1
f

0.15 f2 200

0.1
0
0 200 400 600 800 1000 1200 1400
0.05 H [N]
0 2 4 6 8 10
k
linear until the asymptotic value of k (approximately 4) is reached.
bucket foundations (with a diameter of 293 mm, embedment ra- This linear relationship results in a failure criterion in the radial
tio of 0.5, and low M/DH ratio) subjected to low vertical load in
plane (Fig. 18). Both relationships will fit the measured capacities
loose dry sand. The failure values from those tests can be used to
at low and high impact heights. However, the combined capacity
determine the parameter ␣ in eq. (9) for loose sand. The tension
at low horizontal load appears to decrease abruptly with a slight
capacity of these tests is assumed to correspond to the value of f3W
decrease of k, corresponding to an increase of H in the figure. This
fitted from these tests. The result is shown in Fig. 17 with the
failure criterion does not correspond to any other observation in
corresponding values from the tests presented above. This figure
shows a plausible variation of the inclination factors from Byrne the literature. Thus, the non-linear behaviour observed in Fig. 15 is
et al. (2003) and the present research. The inclination factor is concluded to be a plausible approximation of the f1 and f2.
assumed to tend towards an asymptotic value, as shown previ-
ously. However, the asymptotic value of ␣ is smaller for the tests
Conclusions
of Byrne et al. (2003) than it is in the present research, due to It is common to study failure criteria when seeking to describe
differences in density and strength. At small values of k, the incli- the behavior of bucket foundations subjected to combined loads.
nation factors intersect for the two test series. We have performed a laboratory investigation of the behavior of
If we assume that the variation of ␣ is unaffected by the soil bucket foundations in dense saturated sand subjected to combined
strength, then according to Fig. 17, the variation must be nearly static loads. The results from loading tests with a low vertical load are

Published by NRC Research Press


1258 Can. Geotech. J. Vol. 50, 2013

Fig. 17. Value of inclination factor ␣ from bucket foundation tests caissons. International Journal for Numerical and Analytical Methods in Geo-
with d/D = 0.5. mechanics, 27: 1235–1254. doi:10.1002/nag.319.
Barari, A., and Ibsen, L.B. 2011. Effect of embedment on the vertical bearing
capacity of Bucket foundations in clay. In Proceedings, 2011 Pan-Am CGS
Geotechnical Conference, Toronto, Ont.
Barari, A., and Ibsen, L.B. 2012. Undrained response of bucket foundations to
moment loading. Applied Ocean Research, 36: 12–21. doi:10.1016/j.apor.2012.
01.003.
Barari, A., and Ibsen, L.B. 2013. Vertical capacity of bucket foundations in un-
drained soil. Journal of Civil Engineering and Management. [In press.] doi:10.
3846/13923730.2013.801915.
Bransby, M.F., and Yun, G.-J. 2009. The undrained capacity of skirted strip foun-
dations under combined loading. Géotechnique, 59(2): 115–125. doi:10.1680/
geot.2007.00098.
Butterfield, R. 1981. Another look at gravity platform foundations. In Proceed-
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

ings of Soil Mechanics and Foundation Engineering in Offshore Technology.


CISM, Udine, Italy.
Butterfield, R., and Ticof, J. 1979. Design parameters for granular soils (discus-
sion contribution). In Proceedings of the 7th International Conference on Soil
Mechanics and Foundation Engineering, Brighton. Vol. 4, pp. 259–261.
Byrne, B.W. 2000. Investigation of suction caissons in dense sand. Ph.D. thesis,
University of Oxford.
Fig. 18. Failure criterion from proposed parameter variation
Byrne, B.W., and Houlsby, G.T. 2003. Foundations for offshore wind turbines.
compared with a linear parameter variation. Philosophical Transactions of the Royal Society A, 361: 2909–2930. doi:10.
1098/rsta.2003.1286.
Byrne, B.W., Villalobos, F., Houlsby, G.T., and Martin, C.M. 2003. Laboratory
testing of shallow skirted foundations in sand. In Proceedings, BGA Interna-
tional Conference on Foundations, Dundee, September. pp. 161–173.
Cassidy, M.J. 1999. Non-linear analysis of jack-up structures subjected to random
waves. D.Philos. thesis, Oxford University.
Clukey, E.C., Morrison, M.J., Gamier, J., and Corte, J.F. 1995. The response of
suction caissons in normally consolidated TLP loading conditions. In Proceed-
ings, Offshore Technology Conference, Houston, Tex.
DNV. 1992. Classification notes No. 30.4, Foundations. Det Norske Veritas, Nor-
way.
El-Gharbawy, S., and Olson, R. 2000. Modeling of suction caisson foundations.
For personal use only.

In Proceedings, International Offshore and Polar Engineering Conference.


Vol. 2, pp. 670–677.
El-Gharbawy, S.L., Iskander, M.G., and Olson, R.E. 1997. Application of suction
caisson foundations in the Gulf of Mexico. In Proceedings of the Annual
Offshore Technology Conference. Vol. 1, pp. 531–538.
El-Sherbiny, R.M., Olson, R.E., Gilbert, R.B., and Vanka, S.K. 2005. Capacity of
suction caissons under inclined loading in normally consolidated clay.
In Proceedings, 1st International Symposium on Frontiers in Offshore
Geotechnics (ISFOG). pp. 281–287.
Erbrich, C.T., and Tjelta, T.I. 1999. Installation of bucket foundations and suction
caissons in sand - geotechnical performance. In Proceedings, Offshore Tech-
nology Conference, Houston, Tex.
Feld, T. 2001. Suction buckets, a new innovative foundation concept, applied to
offshore wind turbines. Ph.D. thesis, Aalborg University, Denmark.
Georgiadis, M., and Butterfield, R. 1988. Displacements of footings on sand un-
der eccentric and inclined loads. Canadian Geotechnical Journal, 25(2): 199–
presented. The tests were performed with various embedment, 212. doi:10.1139/t88-024.
V/Vpeak, and load paths (M/DH ratios). Gottardi, G., and Butterfield, R. 1993. On the bearing capacity of surface footings
This study is unique in several key ways. Few studies have ad- on sand under general planar loads. Soils and Foundations, 33(3): 68–79.
doi:10.3208/sandf1972.33.3_68.
dressed the influence of the M/DH and d/D ratios of bucket foun- Gottardi, G., Houlsby, G.T., and Butterfield, R. 1999. Plastic response of circular
dations subjected to low vertical loads. The soil tested in this work footings on sand under general planar loading. Géotechnique, 49(4): 453–
was saturated sand, whereas most previous works in the literature 469. doi:10.1680/geot.1999.49.4.453.
have used dry sand. The use of saturated sand is preferred because Gourvenec, S. 2007. Shape effects on the capacity of rectangular footings under
general loading. Géotechnique, 57(8): 637–646. doi:10.1680/geot.2007.57.8.
the bucket foundation is located offshore.
637.
When we calibrated the failure criterion from the test results, Gourvenec, S., and Barnett, S. 2011. Undrained failure envelope for skirted foun-
we found that the failure parameters depended heavily on the dations under general loading. Géotechnique, 61(3): 263–270. doi:10.1680/
embedment ratio and, to some extent, the load path. Although geot.9.T.027.
previous authors have proposed a linear relationship between the Gourvenec, S., and Randolph, M.F. 2003. Effect of strength non-homogeneity on
the shape of failure envelopes for combined loading of strip and circular
moment capacity and vertical load, this relationship did not fit foundations on clay. Géotechnique, 53(6): 575–586. doi:10.1680/geot.2003.53.
the capacities observed in this study. Therefore, a new expression 6.575.
was suggested to describe the combined capacity as a function of Hansen, J.B. 1970. A revised and extended formula for bearing capacity. Danish
the tensile capacity and inclination factor, the latter of which was Geotechnical Institute Bulletin, 98: 5–11.
observed to be a function of the embedment ratio and the impact Houlsby, G.T., Kelly, R.B., and Byrne, B.W. 2005. The tensile capacity of suction
caissons in sand under rapid loading. In Proceedings of Frontiers in Offshore
height. Geotechnics: ISFOG.
Houlsby, G.T., Kelly, R.B., Huxtable, J., and Byrne, B.W. 2006. Field trials of
References suction caissons in sand for offshore wind turbine foundations. Géotech-
Achmus, M., and Abdel-Rahman, K. 2005. Design of monopile foundations for nique, 56(1): 3–10.
offshore wind energy plants. In Proceedings of the 11th International Colo- Ibsen, L.B. 2008. Implementation of a new foundation concept for offshore wind
quium on Structural and Geotechnical Engineering, Cairo, Egypt. farms. In Proceedings of the 15th Nordic Geotechnical Meeting, Norway.
API. 2000. RP2A: recommended practice for planning, designing and construc- pp. 19–33.
tion fixed offshore platforms. American Petroleum Institute, Washington, Ibsen, L.B., and Barari, A. 2011. Monotonic loading of circular surface footings
D.C. on clay. Journal of Mechanical Science and Technology, 25(12): 3151–3158.
Aubeny, C.P., Han, S.W., and Murff, J.D. 2003. Inclined load capacity of suction doi:10.1007/s12206-011-1001-z.

Published by NRC Research Press


Larsen et al. 1259

Ibsen, L.B., Schakenda, B., and Nielsen, S.A. 2004. Development of the bucket 
ct,a effective cohesion for the tangent parameter at high stress
foundation for offshore wind turbines, a novel principle. Gigawind-Sym. levels
Offshore-Windenergie, Bau- und umwelttechnische Aspekte, Hannover. cu undrained shear strength (kPa)
Ibsen, L.B., Barari, A., and Larsen, K.A. 2012. Modified vertical bearing capacity
D diameter
for circular foundations in sand using reduced friction angle. Ocean Engi-
neering, 47: 1–6. doi:10.1016/j.oceaneng.2012.03.003. Di, Do inner and outer diameter, respectively
Iskander, M., El-Gharbawy, S., and Olson, R. 2002. Performance of suction cais- Dr relative density
sons in sand and clay. Canadian Geotechnical Journal, 39(3): 576–584. doi:10. d depth from soil surface to bottom of bucket foundation (skirt
1139/t02-030. length)
ISO. 2000. Petroleum and natural gas industries: offshore structures. Part 4: E elastic modulus of the soil
Geotechnical and foundation design considerations. ISO19900. International E25 tangent elastic modulus corresponding to 25% strain
Organisation for Standardisation, Geneva. e void ratio
Jacobsen, M. 1970. New oedometer and new triaxial apparatus for firm soil. DGI emax maximum void ratio
Bulletin, 27. Danish Geotechnical Society.
emin minimum void ratio
Kelly, R.B., Byrne, B.W., Houlsby, G.T., and Martin, C.M. 2004. Tensile loading of
model caisson foundations for structures on sand. In Proceedings, ISOPE,
F self-weight of system
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by Virginia Tech Univ Lib on 12/31/15

Toulon. pp. 638–641. fi constant (i = 1, 2, 3)


Luke, A.M., Rauch, A.F., Olson, R.E., and Mecham, E.C. 2005. Components of f1, f2 failure parameters
suction caisson capacity measured in axial pullout tests. Ocean Engineering, f3 parameter representing proportion of soil plug weight mobi-
32: 878–891. doi:10.1016/j.oceaneng.2004.10.007. lized under action of moment loading
Meyerhof, G.G. 1951. The ultimate bearing capacity of foundations. Géotech- Gs grain density
nique, 2: 301–332. doi:10.1680/geot.1951.2.4.301. H horizontal load
Meyerhof, G.G. 1953. The bearing capacity of foundations under eccentric h penetration depth
and inclined loads. In Proceedings of the 3rd International Society for Soil K lateral earth pressure coefficient
Mechanics and Foundation Engineering, Zurich, Switzerland. Vol. 1,
pp. 440–445.
k impact height
Prandtl, L. 1921. Über die Eindringungsfestigkeit ⬃Härte! Plastischer Baustoffe M pure moment load
und die Festigkeit von Schneiden, Z. Angew. Math. Mech, 1: 15–20. m distance over which stresses are reduced outside the founda-
Roscoe, K.H., and Schofield, A.N. 1956. Stability of short pier foundations in tion relative to outer diameter
sand. In Proceedings of Symposium in Cambridge on Plastic Theory of Struc- Nc bearing capacity factor
tures. British Welding Journal, 3(8): 343–354. Nq bearing capacity factor for surcharge
Sukumaran, B., and McCarron, W.O. 1999. Total and effective stress analysis of N␥ bearing capacity factor for self-weight
suction caissons for Gulf of Mexico conditions. In Proceedings, OTRC99 Con- Q drained pullout capacity
ference. Geotechnical Special Publication No. 88. pp. 247–260. Qs frictional resistance on both the inside and outside of the cais-
Taiebat, H.A., and Carter, J.P. 2002. Bearing capacity of strip and circular foun-
dations on undrained clay subjected to eccentric loads. Géotechnique, 52(1):
son wall
For personal use only.

61–64. doi:10.1680/geot.2002.52.1.61. q overburden pressure


Terzaghi, K. 1943. Theoretical soil mechanics. John Wiley and Sons, Inc., New q= deviatoric stress
York. V vertical load
Ukritchon, B., Whittle, A.J., and Sloan, S. 1989. Undrained limit analyses for Vpeak maximum pure vertical load (kN)
combined loading of strip footings on clay. Journal of Geotechnical and Vt vertical tension capacity
Geoenvironmental Engineering, 124(3): 265–276. doi:10.1061/(ASCE)1090- W weight of the soil plug inside the bucket
0241(1998)124:3(265). Wc sum of submerged weights of caisson, templates, and ballast
Vásquez, L.F.G., Maniar, D.R., and Tassoulas, J.L. 2010. Installation and axial ␣ external side resistance
pullout of suction caissons: numerical modeling. Journal of Geotechnical and
Geoenvironmental Engineering, 136(8): 1137–1147. doi:10.1061/(ASCE)GT.1943-
␥ unit weight of the soil
5606.0000321. ␥= effective unit weight of the soil
Vesic, A.S. 1975. Bearing capacity of shallow foundations. In Foundation Engi- ␦ friction angle between the skirt and the soil
neering Handbook. Edited by H.F. Winterkorn and H.Y. Fang. Van Nostrand, ␴1 major effective principal stress
New York. pp. 121–147. ␴3 minor effective principal stress
Wang, D., and Jin, X. 2008. Failure loci of suction caisson foundations under 
combined loading conditions. China Ocean Engineering, 22(3): 455–464. ␴v0 overburden pressure
␸= effective friction angle
List of symbols ␸s triaxial secant friction angle

␸t,a effective friction angle for the tangent parameter at high stress
A foundation area levels
B foundation width ␺ dilation angle
c cohesion
c= effective cohesion

Published by NRC Research Press

View publication stats

You might also like