Beruflich Dokumente
Kultur Dokumente
net/publication/276123623
CITATIONS READS
5 428
2 authors:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Xiaohui Cheng on 06 July 2016.
To cite this article: Hao Wang & Xiaohui Cheng (2015): Undrained Bearing Capacity of Suction Caissons for
Offshore Wind Turbine Foundations by Numerical Limit Analysis, Marine Georesources & Geotechnology, DOI:
10.1080/1064119X.2015.1004766
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Marine Georesources & Geotechnology, 0: 1–13
Copyright # 2015 Taylor & Francis Group, LLC
ISSN: 1064-119X print/1521-0618 online
DOI: 10.1080/1064119X.2015.1004766
Determining the ultimate capacity of suction caissons in response to combined vertical, horizontal, and moment loading is essential
for their design as foundations for offshore wind turbines. However, the method implemented for stability analysis is quite limited.
Numerical limit analysis has an advantage over traditional limit equilibrium methods and nonlinear finite element methods in this
case because upper and lower bounds can be achieved to ensure that the exact ultimate capacity of the caisson falls within the
appropriate range. This article presents theories related to numerical limit analysis. Simulations are conducted for centrifuge model
tests, the findings of which reveal the ability of numerical limit analysis to deal with the inclined pullout capacity of suction cais-
sons. Finally, this article proposes an estimation of the ultimate capacity of a 3.5 MW offshore wind turbine foundation on nor-
mally consolidated clay based on the typical environmental parameters of Bothkennar, Scotland. Undrained failure envelopes and
safety factors are obtained for suction caissons with different embedment ratios. Failure mechanisms, plastic zones, clay stress
distributions, and the influence of the skin friction coefficients of caissons are discussed in detail.
Keywords: limit state design, lower bound, numerical limit analysis, offshore wind turbine, suction caisson foundation, upper
bound
Introduction caisson falls between 0.5 and 1, and is usually smaller than
that of a suction anchor (whose L=D can reach up to 6).
Suction caissons are large hollow cylindrical foundations Two structural types of offshore wind turbines are
that are open at the base and closed at the top and penetrate currently used, the ‘‘monopod’’ and the ‘‘tetrapod’’ (or
into soil with active suction after partial penetration by ‘‘tripod’’) (Figure 1). The diameter of a monopod caisson
deadweight. Depending on the type of load and geometrical is usually 20–25 m (Houlsby et al. 2005b) and that of
conditions, they comprise shallow foundations (suction cais- a tetrapod or tripod caisson is approximately 6–8 m.
sons or suction buckets) or anchors (suction anchors). For The foundation of a tetrapod resists loading mainly via
anchoring systems, suction caissons are usually used in a ‘‘push-pull’’ action. Some caissons suffer a larger
deep-sea operations and suffer uplifts combined with lateral compression and others suffer a smaller compression or even
loads. Many researchers have conducted centrifuge tests of tension in the balance of the moment and self-weight.
suction anchors and piles (Fuglsang and Steensen-Bach Monopod caissons suffer the associative effects of self-
1991; Renzi et al. 1991; Clukey and Morrison 1993; Clukey weight, lateral loads, and overturning moments. In terms
et al. 2004; Zhang et al. 2007), and the foundation type has of analysis, the monopod caisson is much more represen-
been widely used in the North Sea and Gulf of Mexico. The tative than the tetrapod caisson. Therefore, this article
application of suction caissons as shallow foundations, mainly investigates the former.
mainly as offshore wind turbine foundations, is relatively Methods for designing caisson foundations for offshore
new. The suction caissons used for offshore wind turbines wind applications are in their infancy (Houlsby et al.
differ from those used for other purposes in that they are 2005b). The soil-structure interactions and failure modes
installed in much shallower water (depths from near-shore of suction caissons under large overturning moments have
to approximately 40 m) and suffer a much larger moment. not been investigated thoroughly. One key challenge is that
The average embedment ratio (L=D, where L is the wall offshore wind turbine structures are relatively light for their
length and D is the diameter of the caisson cap) of a suction size (for example, a 3.5 MW turbine structure has a mass
of 600 t), despite being subjected to large lateral forces and
Address correspondence to Hao Wang, Department of Civil overturning moments from wind and waves (Byrne and
Engineering, Tsinghua University, Beijing 100084, China. Houlsby 2003). These lateral loads can be as large as 70%
E-mail: wanghao0605@gmail.com of the vertical load.
2 H. Wang and X. Cheng
Research Laboratory (LCPC) and supplied by the Exxon ultimate capacity of a suction caisson foundation. The suc-
Production Research Company) of suction caissons are then tion caisson is placed within the soil with its caisson cap at
compared with numerical limit analysis results to authenti- the mudline. It suffers a vertical load, a horizontal load
cate the accuracy of the latter. Finally, a numerical limit and a moment transferred from the wind turbine structure.
analysis is conducted to estimate the ultimate capacity of a In lower bound analysis, the soil is discretized into
3.5-MW wind turbine foundation on normally consolidated three-node triangular elements. Each node is associated with
clay in Bothkennar, Scotland. Undrained failure envelopes three unknown stresses (rx, ry, sxy) in the Cartesian (x, y)
of combined loading and safety factors are obtained, with frame of reference (Figure 2b). The stress components are
the soil parameters and loading conditions chosen according assumed to vary linearly within each element. In contrast
to the actual situation. to the conventional FE method, each node is unique to
a particular element, so that stress discontinuities may occur
along the shared edges between the elements. The suction
Theories of Numerical Limit Analysis caisson is discretized into two-node beam elements connec-
Background ted by one-node joint elements (of no physical dimension).
Each node is unique to a particular element and is associated
Lysmer (1970) was the first to propose the theory of numeri- with three unknown degrees (two forces and one moment)
cal limit analysis. However, its application was restricted due (Figure 2c). The stresses are assumed to vary linearly along
Downloaded by [Tsinghua University] at 00:26 09 September 2015
to the large computational time required to solve the prob- the structural elements. From the definition, it is also
lems caused by complicated linear programming optimiza- assumed that the beam elements carry only distributed stres-
tion. A revised simplex optimization algorithm introduced ses and that the joint elements carry only point forces. The
by Anderheggen and Knopfel (1972) and used by Bottero soil-structure interface is modeled by the discontinuities
et al. (1980) contributed an important extension to the between the triangular and beam elements (Figure 2d).
method. The efficiency problem was solved by Sloan Lower bound analysis is formulated as a linear program-
(1988b), who introduced a steepest edge active set algorithm ming problem with an objective function subjected to several
into the linear programming. Although the algorithm was constraints (Table 1). The constraints of equalities and
originally developed for finite element formulations of the inequalities are set by the statically admissible stress field
lower bound theorem, the scheme has also proven to be very subjected to equilibrium (within soil and structural elements
efficient for upper bound analysis (Sloan and Kleeman 1995). and along discontinuities and soil structure interfaces), stress
As efficiency improved, simple strip footing cases were boundary conditions and the yield criterion. A linearized
discussed for both the lower (Sloan 1988a) and upper (Sloan Tresca yield criterion is applied in this paper (Figure 2e),
and Kleeman 1995) bounds. Ukritchon et al. (1998; 2003) offering the convenience of linear programming and giving
extended the application of the method to a combined the analysis a rigorous lower (interior polygon) or upper
loading of strip footings on clay and the undrained stability (exterior polygon) bound. A 24-sided polygon is used for
of braced excavations. both the upper and lower bounds with the consideration
Salgado et al. (2004) recently proposed a three- of accuracy and efficiency. Sloan (1988a) describes the
dimensional type of numerical limit analysis. A robust constraint formulations in more detail.
nonlinear algorithm was used for the lower bound and a The objective function of the lower bound is to maximize
quasi-Newton strategy for the upper bound (Lyamin and the driving force that leads to failure under the constraints of
Sloan 2002). The method has proven to be accurate and a statically admissible stress field. In suction caisson analysis,
efficient for estimating the shape factors of square and rec- the critical driving forces are the lateral loads and moments
tangular footings, which could not be derived in the past sent to the caissons by the supporting structures. The loads
due to the lack of three-dimensional computations. The are a combination of wind and waves. Finally, the linear
ability of three-dimensional analysis to deal with complex soil programming problem is solved using the active set
behavior and loading conditions remains quite limited and algorithm (Sloan 1988b).
restricted to footings and soil of uniform undrained strength.
Due to the complexity of suction caisson foundations for
offshore wind turbines, this article adopts two-dimensional
numerical limit analysis as its computational method. This type Upper Bound Analysis
of limit analysis is easier to visualize than the three-dimensional The type of soil and structure discretization used in lower
type and requires less processing time. In addition, researchers bound analysis is also used in upper bound analysis
have found that the normalized combined loading capacity of (Figure 2b). Each node of a beam or a triangular element
plane-strain and circular surface foundations varies insignifi- has two unknown velocities in the X and Y directions.
cantly (Bransby and Yun 2009). A brief introduction of the A joint element has three unknown velocities (an additional
method framework is provided as follows. angular velocity w enables the formation of plastic hinges
along the structure, as shown in Figure 2c). The velocity field
varies linearly within the elements. Due to the uniqueness of
Lower Bound Analysis
each node to a particular element, velocity discontinuities
Figure 2a shows a brief scheme of two-dimensional plane- are allowed along the shared edges. The discontinuities
strain numerical limit analysis used for analyzing the between the soil and caisson shown in Figure 2d enable slip
4 H. Wang and X. Cheng
Downloaded by [Tsinghua University] at 00:26 09 September 2015
Fig. 2. Problem discretization for caisson foundations and summary of plane-strain elements in numerical limit analysis: (a) prob-
lem discretization; (b) soil elements of numerical limit analysis; (c) structural elements of numerical limit analysis; (d) modeling for
soil-structure interface; and (e) polygonal representation of Tresca yield criterion.
shear strength. The contour of the plastic zone in the foundation, and that from wind is 1 MN and acts at 90 m
front (the positive direction is defined by the horizontal above the foundation. Therefore, the total lateral load can
component of the pullout force) is mainly vertical and that be represented by a 4 MN force acting at 30 m above the
Downloaded by [Tsinghua University] at 00:26 09 September 2015
at the back extends in an angle of approximately 45 . This foundation, which leads to a 120 MNm overturning
is not a plastic zone of symmetry, as other researchers have moment. The vertical load is the self-weight of the wind
suggested (e.g., Murff and Hamilton 1993), indicating that turbine, which is 6 MN in this case.
the former upper bound plastic limit solutions may result The water depth is assumed to be 20 m. The diameter of
in errors when estimating the failure mechanism. the suction caisson is usually 20–25 m (Houlsby et al.
Lower bound analysis provides the distributions of 2005b). In this section, 20 m is chosen as the diameter for
major and minor principal stress (Figure 5a; 5b) and the all of the caissons. Embedment ratios (L=D) of 0.5, 0.75,
direction of the major principal stress to the vertical d and 1.0 are discussed, indicating lengths of 10, 15, and
(Figure 5c). Stress concentration develops near the tip of 20 m for the caisson walls.
the front and back walls (mainly at the back wall), which The soft clay at Bothkennar is unusual in that it is
may be the weak region of the clay layer under this typical remarkably uniform, with its engineering properties being
loading condition (see Figure 5a; 5b). The uplift loading consistent across the site and increasing linearly with the
causes a vertical release of stress in the soil elements depth (Nash et al. 1992). According to the in situ vane tests
directly beneath the caisson. At the failure point, the described by Nash et al. (1992), the average vane strength
vertical stress component of these soil elements acts as profile is given by
the minor principal stress rather than as the major principal
su ðkPaÞ ¼ 14 þ 2:3 zðmÞ ð3Þ
stress (see Figure 5c). A zone of passive failure occurs
outside the front wall (with reference to Figure 4a), with where su is the undrained shear strength of the clay at differ-
a degree of 90 of the major principal stress to the vertical. ent depths. The buoyant unit weight of clay is 7.0 kN=m3.
In the back wall, this angle only occurs around the tip.
The rotational angle of the principal stress suggests that
further analysis considering the anisotropy of the clay
may lead to more accurate results.
No tension cut-offs are assumed for the undrained strength moment capacity and vice versa. As expected, a larger
of the clay. embedment ratio leads to a larger envelope. The pure lateral
Lower bound analysis uses a high density of elements capacity H0 (when the moment is zero) of L=D ¼ 1.0 is 29%
around the caisson, especially at the caisson tips, where larger than that of L=D ¼ 0.75 and 77% larger than that of
stress concentration may occur. Extension elements are L=D ¼ 0.5, and the pure moment capacity M0 (when the
applied to ensure the statically admissible stress field in half lateral load is zero) of L=D ¼ 1.0 is 55% higher than that
space (Ukritchon et al. 2003). A regular mesh is used for the of L=D ¼ 0.75 and 172% higher than that of L=D ¼ 0.5.
upper bound analysis. These results indicate that elongation of the caisson walls
The analysis assumes that no gap occurs between the cap has a more obvious effect on resisting moments than on
and underlying soil or along the soil-structure interfaces. resisting lateral loads.
The skin friction coefficient a (mainly relating to the Figure 8 plots the normalized failure envelopes of the
installation procedure and material properties) is assumed to means of the upper and lower bound results, where M0 is
be 1.0 in the following sections, but is discussed in detail later. the pure moment capacity and H0 is the pure horizontal
capacity. The results are compared with the FEM results
produced by Bransby and Yun (2009) (FEM-1) and Hung
Ultimate Bearing Capacity and Failure Envelope and Kim (2014) (FEM-2). In their analyses, the soil was
The failure envelope approach has recently been adopted to modeled as an elastic perfectly plastic material and the foun-
Downloaded by [Tsinghua University] at 00:26 09 September 2015
express the capacity of shallow foundations (e.g., Bransby dations were modeled as solid, rigid bodies. The two FEM
and Yun 2009; Achmus et al. 2013; Hung and Kim 2014). calculations assume that V ¼ 0. Although the strength para-
The loads contributing to failure are plotted in the meters of the three analyses are different, the comparison is
horizontal-moment (H-M) load space (if the vertical load worthwhile. According to Figure 8, the normalized failure
(V) is assumed to be constant), thus forming a failure envel- envelopes vary according to the different embedment ratios.
ope. Load states within the failure envelope do not cause The gaps between the envelopes revealed by the numerical
failure, while those on the envelope do. Therefore, a simple limit analysis are relatively small compared with the FEM
design method involves comparing the applied load with results. Although the shapes of L=D ¼ 1.0 determined by
failure envelopes in the load space to ensure that the chosen FEM and the numerical limit analysis are quite similar,
foundation design has an adequate safety margin. Numerical those of L=D ¼ 0.5 show some differences, perhaps due
limit analysis is used to investigate the failure envelopes of to the difference in the vertical load and soil conditions.
suction caissons with different embedment ratios.
Given the geological situation described in the previous
section and the lateral load H, self-weight V and overturning Factor of Safety
moment M acting at the center of the caisson cap, the failure The factor of safety (FOS) is the ratio of the ultimate
envelopes of suction caissons in H-M space are obtained and capacity to the applied load in relation to the same
plotted in Figure 7. As defined in Figure 2a, both H and M self-weight and M=H ratio. The applied load of the wind tur-
are positive in order to achieve the combined extreme state. bine is supposed to be a 4-MN lateral force added to the pole
Figure 7 shows that the failure envelopes vary signifi- 30 m above the mudline, as shown in Figure 6 and by the star
cantly according to different embedment ratios. In all of point at (4, 120) in Figure 7. The ultimate capacity is reached
the cases, the lateral load increment leads to a reduction in
Fig. 7. H-M failure envelopes of caisson foundations with Fig. 8. Normalized failure envelopes of caisson foundations
different embedment ratios. in H-M space.
Undrained Bearing Capacity of Suction Caissons 9
Table 3. Factor of safety for suction caisson foundation Figure 9 plots the specific failure mechanisms. If the moment
dominates, then the failure mechanism is a rigid body rotation
FOS LB Mean UB Bound Errora with its center located at a certain depth under the cap (Points
A and D). If the lateral load dominates or the lateral load and
L=D ¼ 1.0 4.56 4.95 5.35 7.94%
moment combine, the failure mechanism involves both a rigid
L=D ¼ 0.75 3.16 3.42 3.68 7.59%
L=D ¼ 0.5 2.01 2.14 2.28 6.50%
body rotation and a lateral sliding (Points B, C, E, and F).
a
For the points on a failure envelope of a certain embedment
Calculation of bound error: 0.5 (UB-LB)=Mean..
ratio, as H increases (e.g., from A to C or from D to F): (1)
more of the ground surface in front of the caisson heaves
by maximizing the external load under the constant while that behind the caisson settles; and (2) the depth of
self-weight and M=H ratio, represented by the intersections the external load’s influence becomes shallower, suggesting
between the dashed line and the failure envelopes of each that the failure mechanism tends to be rather horizontal than
embedment ratio in Figure 7. vertical.
Table 3 indicates that, in all three of the cases, the system
is able to resist the extreme load with an FOS larger than 2.
The FOS increases with the embedment ratio. Meanwhile, Plastic Zone
Downloaded by [Tsinghua University] at 00:26 09 September 2015
the effect of an increment from 0.75 to 1.0 is not as obvious Determining which part of the clay layer will fail is always a
as that of an increment from 0.5 to 0.75. Engineers and major issue in a stability analysis. A plastic zone represents
researchers must set the expected FOS value before the clay failure in the upper bound analysis. Figure 9 clearly
design to decide the embedment ratio of suction caisson shows that caissons with larger embedment ratios take great-
foundations in relation to the performance cost ratio. In er advantage of deep-level clay, resulting in a deeper plastic
all of the cases, the bound error is within 8%, suggesting zone and a larger bearing capacity. The magnitude of the lat-
that the mean of the upper and lower bounds is a good eral load mainly affects the plastic zone horizontally and
estimation of the true ultimate bearing capacity. that of the overturning moment affects the zone vertically
(compare A with C or D with F).
When moment loading prevails (Points A and D), the
Failure Mechanism
plastic zone is in agreement with the plastic limit approach
The upper bound analysis shows that different failure proposed by Murff and Hamilton (1993). It assumes a sur-
mechanisms prevail at different points on the failure envelope. face failure wedge, a plane-strain flow-around zone and a
Fig. 9. Upper bound mesh and deformation patterns for caisson foundations of L=D ¼ 1.0 and L=D ¼ 0.5 at different ultimate
states (Points A, B, C, and D, E, F in Figure 7).
10 H. Wang and X. Cheng
Fig. 12. Lower bound mesh and principal stress distributions for caisson foundations of L=D ¼ 1.0 at different ultimate states
(Points R, S, T in Figure 7).
the excess pore water pressure dissipates. Experience has within 15%. The reduction of the skin friction weakens the
shown that the affected area is mainly distributed along bond between the caisson and clay layer, causing more
the caisson walls and that the area as a whole suffers a plasticity within the caisson and a weaker influence on the
reduction in undrained shear strength. Renzi et al. (1991) deep layers (Figure 14). Mana et al. (2010) proved via FEM
and Renzi and Maggioni (1994) assumed that the influence that reductions in the vertical capacity of a suction caisson
of the changed undrained shear strength of the area could
be represented by a reduction of skin friction along the walls.
Fuglsang and Steensen-Bach (1991) and Renzi et al. (1991)
suggested that the skin friction could be calculated by
multiplying the average undrained
shear strength by the skin
friction coefficient a: f ¼ Su a, where f is the decreased shear
strength of the side walls (both sides), 0 a 1, considering
both the influence of the installation and the material
properties along the soil-structure interfaces. Therefore, by
using the same geological condition described earlier and
a constant ratio of M=H ¼ 30, the ultimate capacity (P) under
different values of a is obtained. a usually stands between 0.7
and 1.0. However, to fully observe the influence of a on the
bearing capacity, a range from 0 to 1 is analyzed here.
Figure 13 shows the results of the ultimate capacity
under different values of a compared with a ¼ 1. The figure
indicates that the skin friction coefficient has a significant
influence on the ultimate capacity of the caisson and
that the capacity drops faster as a decreases in all three of
the cases. In common situations, where a is between 0.7 and Fig. 13. Normalized ultimate capacity of caisson foundations
1.0, the decreased amplitude of the ultimate capacity remains with different skin friction coefficients.
12 H. Wang and X. Cheng
Conclusions Part of this article was finished while the first author was
visiting the Massachusetts Institute of Technology with the
This article applied numerical limit analysis to investigate support of Huang Yanfang=Xinhe Educational Funds.
the undrained combined loading capacity of the suction
caisson foundations of offshore wind turbines. Simulations References
were conducted for normally consolidated clay, in which
Achmus, M., C. T. Akdag, and K. Thieken. 2013. Load-bearing
the soil strength increased linearly with depth. The following behavior of suction bucketfoundations in sand. Applied Ocean
main conclusions can be drawn: Research 43: 157–165.
. Based on the back-calculation of the inclined pullout Anderheggen, E. and H. Knopfel. 1972. Finite element limit analysis
using linear programming. International Journal of Solids and
capacity of the EPR centrifuge tests, the numerical limit
Structures 8: 1413–1431.
analysis estimated the test results well, with an error bound Aubeny, C. P., S. W. Han, and J. D. Murff. 2003. Inclined load
of 5.75%. An asymmetric plastic zone was obtained, capacity of suction caissons. International Journal for Numerical
indicating that the earlier upper bound limit approaches and Analytical Methods in Geomechanics 27: 1235–1254.
using symmetric plastic zones might have contained errors. Barari, A. and L. B. Ibsen. 2012. Undrained response of bucket
. The analysis proved the ability of suction caissons to func- foundations to moment loading. Applied Ocean Research 36: 12–21.
tion as foundations for offshore wind turbines from a Bottero, A., R. Negre, J. Pastor, and S. Turgeman. 1980. Finite
element method and limit analysis theory for soil mechanics
bearing capacity viewpoint. For a typical 3.5 MW offshore problems. Computer Methods in Applied Mechanics and Engineer-
wind turbine at Bothkennar, Scotland, the safety factors ing 22: 131–149.
of caissons with a diameter of 20 m and embedment ratios Bransby, M. F. and M. F. Randolph. 1998. Combined loading of
of 0.5, 0.75, and 1.0 were 2.14, 3.42, and 4.95, respectively. skirted foundations. Géotechnique 48(5): 637–655.
Undrained Bearing Capacity of Suction Caissons 13
Bransby, M. F. and G. J. Yun. 2009. The undrained capacity of skirted Frontiers in Offshore Geotechnics II, ISFOG 2010. Gourvenec, S.
strip foundations under combined loading. Géotechnique 59(2): and D. White (eds.), 433–438. London: Taylor and Francis Group.
115–125. Murff, J. D. and J. M. Hamilton. 1993. P-ultimate for undrained analy-
Byrne, B. W. and G. T. Houlsby. 2003. Foundations for offshore wind sis of laterally loaded piles. Journal of Geotechnical Engineering
turbines. Philosophical Transactions. Series A, Mathematical, 119(1): 91–107.
Physical, and Engineering Sciences 361: 2909–2930. Nash, D. F. T., J. J. M. Powell, and I. M. Lloyd. 1992. Initial investiga-
Cassidy, M. J., B. W. Byrne, and G. T. Houlsby. 2002. Modelling the tions of the soft clay test site at Bothkennar. Géotechnique 42(2):
behaviour of circular footings under combined loading on loose 163–181.
carbonate sand. Géotechnique 52(10): 705–712. Renzi, R. and W. Maggioni. 1994. Modeling the behavior of skirt piles.
Clukey, E. C., C. P. Aubeny, and J. D. Murff. 2004. Comparison of Proceedings of the 7th International Conference on the Behavior of
analytical and centrifuge model tests for suction caissons subjected Offshore Structures, Cambridge, MA, 77–85.
to combined loads. Journal of Offshore Mechanics and Arctic Renzi, R., W. Maggioni, and F. Smits. 1991. A centrifugal study on
Engineering 126(4): 364–367. the behavior of suction piles. Proceedings of the International
Clukey, E. C. and M. J. Morrison. 1993. A centrifuge and analytical Conference on Centrifuge Modelling, Boulder, CO, 169–176.
study to evaluate suction caissons for TLP applications in the Gulf Salgado, R., A. V. Lyamin, S. W. Sloan, and H. S. Yu. 2004. Two- and
of Mexico. Design and Performance of Deep Foundations: Piles and three-dimensional bearing capacity of foundations in clay.
Piers in Soil and Soft Rock, ASCE, NY: 141–156. Géotechnique 57(8): 647–662.
Fuglsang, L. D. and J. O. Steensen-Bach. 1991. Breakout resistance of Sloan, S. W. 1988a. Lower bound limit analysis using finite elements
suction piles in clay. Proceedings of the International Conference on and linear programming. International Journal for Numerical and
Downloaded by [Tsinghua University] at 00:26 09 September 2015
Centrifuge Modelling, Boulder, CO, 153–159. Analytical Methods in Geomechanics 12(1): 61–77.
Houlsby, G. T., L. B. Ibsen, and B. W. Byrne. 2005a. Suction caissons Sloan, S. W. 1988b. A steepest edge active set algorithm for solving
for wind turbines. Frontiers in Offshore Geotechnics, ISFOG 2005: sparse linear programming problems. International Journal for
Proceedings of the 1st International Symposium on Frontiers in Numerical Methods in Engineering 26(12): 2671–2685.
Offshore Geotechnics: 75–93. Sloan, S. W. 1989. Upper bound limit analysis using finite elements and
Houlsby, G. T., B. W. Byrne, J. Huxtable, and R. B. Kelly. 2005b. linear programming. International Journal for Numerical and Ana-
Field trials of suction caissons in clay for offshore wind turbine lytical Methods in Geomechanics 13: 263–282.
foundations. Géotechnique 55(4): 287–296. Sloan, S. W. and P. W. Kleeman. 1995. Upper bound limit analysis
Hung, L. C. and S. R. Kim. 2012. Evaluation of vertical and horizontal using discontinuous velocity fields. Computer Methods in Applied
bearing capacities of bucket foundations in clay. Ocean Engineering Mechanics and Engineering 127: 293–314.
52: 75–82. Tani, K. and W. H. Craig. 1995. Bearing capacity of circular founda-
Hung, L. C. and S. R. Kim. 2014. Evaluation of undrained bearing tions on soft clay of strength increasing with depth. Soils and
capacities of bucket foundations under combined loads. Marine Foundations 35(4): 21–35.
Georesources & Geotechnology 32(1): 76–92. Ukritchon, B., A. J. Whittle, and S. W. Sloan. 1998. Undrained limit
Ibsen, L. B., A. Barari, and K. A. Larsen. 2014a. Adaptive plasticity analyses for combined loading of strip footings on clay. Journal
model for bucket foundations. Journal of Engineering Mechanics of Geotechnical and Geoenvironmental Engineering 124(3): 265–276.
140(2): 361–373. Ukritchon, B., A. J. Whittle, and S. W. Sloan. 2003. Undrained stab-
Ibsen, L. B., K. A. Larsen, and A. Barari. 2014b. Calibration of failure ility of braced excavation in clay. Journal of Geotechnical and
criteria for bucket foundations on drained sand under general Geoenvironmental Engineering 129(8): 738–755.
loading. Journal of Geotechnical and Geoenvironmental Engineering Whittle, A. J., J. T. Germaine, D. F. Cauble, M. A. Geer, and B.
140(7): 04014033. Ukritchon. 1996. Behavior of Suction Caisson Foundations Final
Larsen, K. A., L. B. Ibsen, and A. Barari. 2013. Modified expression Report. Boston: Massachusetts Institute of Technology.
for the failure criterion of bucket foundations subjected to com- Yuan, Y. and A. J. Whittle. 2013. Evaluation and prediction of 17th
bined loading. Canadian Geotechnical Journal 50(12): 1250–1259. Street Canal I-wall stability using numerical limit analyses. Journal
Lesny, K. 2010. Foundations for Offshore Wind Turbines: Tools for of Geotechnical and Geoenvironmental Engineering 139(6):
Planning and Design. Essen, Germany: VGE Verlag GmbH. 841–852.
Liu, M., M. Yang, and H. Wang. 2014. Bearing behavior of Zhang, J. H., Z. Y. Chen, and F. Li. 2010. Three dimensional limit
wide-shallow bucket foundation for offshore wind turbines in analysis of suction bucket foundations. Ocean Engineering 37:
drained silty sand. Ocean Engineering 82: 169–179. 790–799.
Lyamin, A. V. and S. W. Sloan. 2002. Lower bound limit analysis using Zhang, J. H., L. M. Zhang, and X. B. Lu. 2007. Centrifuge
non-linear programming. International Journal for Numerical modeling of suction bucket foundations for platforms under
Methods in Engineering 55: 573–611. ice-sheet-induced cyclic lateral loadings. Ocean Engineering
Lysmer, J. 1970. Limit analysis of plane problems in soil mechanics. 34: 1069–1079.
Journal of Soil Mechanics and Foundations Division 96: 1311–1334. Zhu, B., D. Q. Kong, R. P. Chen, L. G. Kong, and Y. M. Chen. 2011.
Mana, D. S. K., S. Gourvenec, and M. F. Randolph. 2010. A numerical Installation and lateral loading tests of suction caissons in silt.
study of the vertical bearing capacity of skirted foundations. Canadian Geotechnical Journal 48(7): 1070–1084.