Professional Documents
Culture Documents
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/281135832
READS
244
1 AUTHOR:
Cristina Vulpe
University of Western Australia
8 PUBLICATIONS 22 CITATIONS
SEE PROFILE
All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Cristina Vulpe
letting you access and read them immediately. Retrieved on: 19 March 2016
Vulpe, C. Gotechnique [http://dx.doi.org/10.1680=geot.14.P.200]
The response of skirted circular foundations with rough and smooth soilskirt interface to combined
loading is investigated. The shape and size of the failure envelopes of the skirted circular foundations
under a practical range of embedment ratio, soil strength heterogeneity, soilskirt interface and level
of vertical mobilisation are compared to those of solid embedded circular foundations and skirted
strip foundations. The results show that both the foundation geometry and soil plug inside the
skirt compartment significantly influence the uniaxial capacity and the shape and size of the failure
envelopes. Approximating expressions for solid embedded foundations do not capture the shifting
eccentricity of the failure envelopes of the skirted circular foundations. A new approximating solution
for describing the failure envelope as a function of embedment ratio, soil strength heterogeneity and
soilskirt interface is proposed. Scaling factors accounting for the decreasing uniaxial capacity of
skirted foundations with smooth soilskirt interface are also defined.
1
2 DESIGN METHOD FOR THE UNDRAINED CAPACITY OF SKIRTED FUONDATIONS
soilskirt roughness interface and subjected to 3D loading D
were carried out.
The purpose of the work presented in the current study is
10 D
soil plug inside the skirts compartment can overestimate
capacity
(b) to investigate the effect of soilstructure roughness
interface (i.e. soilskirt interface) on the bearing capa-
city of skirted circular foundations by showing that the
interface roughness represents an important factor in
10 D
the shape and size of the failure envelope
(c) to define a simple algebraic expression for the failure
envelopes of skirted circular foundations for a practical Fig. 1. Example of FE mesh (d=D = 050)
range of embedment ratios, soil strength heterogeneities
and soilskirt interfaces.
& Randolph, 2007) and particle image velocimetry analyses
Based on the outcomes resulting from points (a) and (b), (Mana et al., 2012). Gaps along the soilskirt interface are
namely that both the shape and size of the failure envelopes expected to form and remain open on overconsolidated soil
are misrepresented by available solutions for other type deposits (Britto & Kusakabe, 1982; Supachawarote et al.,
of foundations, it is necessary to explicitly construct failure 2005; Mana et al., 2013). Since only weak normally con-
envelopes (point (c)) for skirted circular foundations. solidated soil was considered in this study, the effect of gaps
Finally, the combined capacity of a skirted strip foun- along the external side was not investigated.
dation with rough soilskirt interface (Gourvenec & Barnett, The fully bonded (rough) interface leads to an optimistic
2011), a solid embedded circular foundation with rough soil assessment of failure. Provided that full adhesion is achieved
structure interface (Vulpe et al., 2014) and skirted circular inside the skirt compartment, the ultimate limit state
foundations with both rough and smooth soilskirt interface response under fully smooth soilskirt external side interface
(current study) are compared where the same embedment conditions is considered conservative. In reality, a level of
ratio, soil strength profile and vertical loading are applied. frictional contact is expected between the soil and the outer
side of the skirts, as demonstrated by experimental investi-
gations of skirted foundations on clay (House & Randolph,
FINITE-ELEMENT MODEL 2001; Andersen & Jostad, 2002; Chen & Randolph, 2007;
All 3D small-strain FE analyses modelling skirted circular Mana et al., 2012, 2013). It is important to note that the
foundations, carried out in this study, were performed using results in the current study would prove unconservative for
Abaqus commercial FE computer software (Dassault skirted foundations with soilskirt internal wall of inter-
Systmes, 2012). mediate roughness or fully smooth.
The undrained shear strength of the normally consolidated
(NC) soil considered in this study was assumed as either
Model geometry, mesh and material parameters uniform or linearly increasing with depth and defined by
Skirted circular foundations with embedment depth (d ) su sum kz 1
to foundation diameter (D) ratios of d=D [0, 010, 025,
050] and skirt thickness (t) to foundation diameter ratio where sum is the shear strength at the mudline and k is the
t=D 0005 for all embedment ratios were considered. Three- shear strength gradient with depth z (Fig. 2). The degree of
dimensional FE meshes were used to model the circular soil strength heterogeneity can be expressed in terms of a
foundations. Owing to symmetry along the vertical centreline
of the foundation, only half of the problem was modelled. A
typical representation of the FE mesh is illustrated in Fig. 1. D
The mesh boundary extends ten times the foundation dia-
meter both horizontally and vertically from the centreline of
the foundation to ensure foundation response independent
of boundary effects. The vertical boundaries prohibited sum su0 su
displacement around the mesh circumference and the base
of the mesh was fixed in all three directions.
The skirted foundations were represented as rigid bodies d
with a single load reference point (RP) located at skirt tip
level along the centreline of the foundation. Since small- RP
strain FE analyses were considered, the installation process of
the skirted foundations was not modelled the foundations 1
were considered wished-in-place.
The soilskirt roughness interface was defined as follows: k
the soilskirt internal side interface was prescribed as fully
bonded (accounting for undrained response, i.e. suction, to
uplift) while for the soilskirt outer side interface either a
z
fully bonded (i.e. rough in shear and no separation allowed)
or a fully smooth (no adhesion) interface was considered.
The soilskirt internal side interface was assumed fully Fig. 2. Definition of notation for foundation geometry, reference point
bonded based on centrifuge experimental evidence (Chen and soil shear strength profile
VULPE 3
dimensionless index as Load paths
All pure vertical (Vult), horizontal (Hult) and moment
kD
2 (Mult) capacities (i.e. obtained in the absence of other
sum load components, for example V for H M 0, and referred
rendering the FE results applicable for a variety of NC soil to herein as uniaxial capacities) were performed with
profiles with strength linearly increasing with depth and displacement-controlled probes applied to the RP until
skirted foundations of varied absolute dimensions. Values of failure was reached manifested through increasing dis-
0 (uniform shear strength with depth), 6, 20, 60 and 100 placement under constant load. Only small displacement
(essentially NC) were considered. excursions were necessary to reach the ultimate limit state.
The NC soil was modelled as a linear elastic perfectly General combined VHM loading was achieved by apply-
plastic material obeying Tresca failure criterion (i.e. ing a vertical load as a direct force, after which a series of
max su). The soil was prescribed an undrained Youngs constant-ratio displacement probes of translation (u) and
modulus linearly increasing with depth with constant rotation () were applied to the RP. The applied vertical load
Eu=su 500, Poisson ratio 0499 and effective unit level was defined as a proportion of the uniaxial vertical
weight 6 kN=m3, which are realistic values for a soft capacity (Vult) of either rough or smooth skirted foundation,
marine clay. The submerged part of the foundation system, respectively. The vertical load level is described by v V=Vult,
namely the skirts, was prescribed identical effective unit where v took values of 0, 050 or 075. An example of a failure
weight as the soil in order to obtain the net capacity of the envelope is presented in Fig. 3. In general, between ten
foundation independent of buoyancy effects. and 15 probe tests were required to satisfactorily construct
a failure envelope for each combination of level of vertical
load, soil strength heterogeneity index, embedment ratio and
soilskirt interface.
Envelope
Sign convention and notation
Constant displacement ratio probe Sign conventions for displacements and loads presented in
Mmax u
15
D
= 05 this study follow the Butterfield et al. (1997) recommen-
dations, as shown in Fig. 4. The notations adopted herein are
u summarised in Table 1. Maximum values of horizontal and
= 025
D
10 moment capacities (denoted Hmax and Mmax, respectively)
do not coincide with pure loading as a result of the positive
M/ADsu0
u
= 05 u
D
D
= 10 effect of cross-coupling between horizontal and moment
degrees of freedom. Bearing capacity factors for each load-
05 ing direction are derived from uniaxial bearing capacities
u
= 10 u with respect to the undrained shear strength at skirt tip level
D = 30
D
su0 sum kd.
Hmax
0
15 10 05 0 05 10 15
Validation
H/Asu0
The accuracy of the FE meshes was compared against
available theoretical solutions of ultimate limit states under
Fig. 3. Example of failure envelope using probe tests from FE pure vertical (Martin & Randolph, 2001; Martin, 2003),
analyses (d=D = 0, = 6, V = 0)
horizontal (H=Asu 1) and moment loading (Randolph &
Puzrin, 2003) for the surface foundations because of a lack
of exact solutions for skirted circular foundations. For the
range of soil heterogeneities studied, the FE results for pure
M vertical capacity agreed to within 2% of the theoretically
H V
RP exact solutions (Martin, 2003). The pure horizontal capacity
w RP resulted in good agreement for low-strength heterogeneity
indices but was increasingly overestimated (up to 10% for
u 100) with increasing soil strength heterogeneity as
ultimate limit state coincided with a thin layer of elements
Fig. 4. Sign convention and notation confining the failure mechanism. Pure moment capacity was
Displacement w u
Load V H M
Pure uniaxial capacity Vult Hult Mult
Maximum capacity Hmax Mmax
Normalised load v V=Vult h H=Hult m M=Mult
Bearing capacity factor for a skirted foundation NcV Vult=Asu0 NcH Hult=Asu0 NcM Mult=ADsu0
with rough soilskirt interface
smooth smooth smooth
Bearing capacity factor for a skirted foundation NcV NcH NcM
with smooth soilskirt interface
Scaling factor v NcV
smooth
=NcV h NcH
smooth
=NcH m NcM
smooth
=NcM
low embedment ratios and high soil heterogeneity, and the Skirted foundation, smooth interface
resulting horizontal bearing capacity is only slightly lower (current study)
20
than for an embedded circular foundation around 8%
smaller for d=D 01 and 100.
Smooth skirted foundations report lower horizontal 15
capacities compared to the rough skirted foundations for
all embedment ratios and soil heterogeneities investigated.
10
Lowest horizontal capacities are obtained for smooth skirted
foundations with d=D 01 on soil with . 0. The difference
in horizontal bearing capacity between smooth and rough 05
skirted foundations decreases with increasing embedment 0 01 02 03 04 05
ratio. Fig. 7 allows explanation for this: the kinematic mech- Embedment ratio, d/D
anism at failure for d=D 010 involves an inverted scoop (c)
reaching higher into the skirt compartment with increasing
soil heterogeneity, while longer skirts (d=D . 01) and low Fig. 5. Comparison of uniaxial bearing capacity factors between
either reduce or inhibit this effect with minimal soil failure embedded foundations (with rough soilstructure interface) and
penetration within the skirts for d=D 025 and , 20. skirted foundations (with rough and smooth soilskirt interface) for
(a) pure vertical loading, (b) pure horizontal loading and (c) pure
Figure 5(c) summarises the uniaxial moment capacities
moment loading
for solid embedded circular foundations and rough and fully
smooth skirted circular foundations. All skirted foundations
report lower moment capacities than the solid embedded with a centre of rotation above the foundation is observed for
foundations and it is not conservative to assume a solid plug the solid foundation. The fully smooth skirt outer side does
inside the skirt compartment. The results are also investi- not contribute to load shedding, resulting in a significant
gated from a kinematics to failure perspective in Fig. 6. The part of the loading being transferred to skirt tip level. This
difference in failure mechanisms is obvious an internal leads to a more pronounced internal scoop mechanism for
scoop mechanism with a rotation point beneath the foun- the smooth skirted foundation, which reports the lowest
dation develops for the skirted foundation (failure is pushed moment capacities (up to 35% lower than the rough skirted
upwards in shallower weaker soil), while an outward scoop foundation for d=D 050 and 20).
VULPE 5
d/D = 050 Skirted foundation Embedded foundation Skirted foundation
= 20 Rough interface Rough interface Smooth interface
Vult
Hult
Mult
Fig. 6. Comparison of the failure mechanisms (shown as total displacement vectors) under uniaxial loading between embedded foundations (with
rough soilskirt interface) and skirted foundations (with rough and smooth soilskirt interface)
Hult =0 = 20 = 100
d/D = 010
d/D = 025
d/D = 050
Fig. 7. Comparison of failure mechanisms (shown as total displacement vectors) under uniaxial horizontal loading (Hult) for discrete levels of
embedment ratio d=D = 010, 025, 050 and soil strength heterogeneity index = 0, 20, 100 for the fully smooth skirted foundation
The combined VHM capacity of skirted foundations is Skirted foundation, rough interface (current study)
next compared to that of solid embedded solutions in Fig. 8. Embedded foundation, rough interface (Vulpe et al., 2014)
As noted by Bransby & Yun (2009), both the shape and size
of the failure envelope for a skirted foundation differ from = 60 15 d/D = 010
that of an embedded foundation, especially for foundation V/Vult = 0 d/D = 050
systems with low embedment ratios. All failure mechanisms
contain a rotational () component to failure (except for the
abscissa apex point) and moment capacity has been shown to 10
m/m*
and
0
NcMd=D; 0 01 02 03 04 05
dcM for moment capacity 5 Embedment ratio, d/D
NcMd=D0;
(b)
where NcV(d=D,), NcH(d=D,) and NcM(d=D,) are the capacity
factors for any given embedment ratio d=D and soil strength 25 Rough interface
heterogeneity index , and generically called NcV, NcH and
NcM (see Table 1). The depth factors for rough skirted = 0, 6, 20, 60, 100
20
foundations with embedment ratios 0 d=D 05 and
0 100 are illustrated in Fig. 9 and also summarised
Depth factor, dcM
0 0 1 1 1 10
6 1 1 1
5
20 1 1 1
60 1 1 1 0
100 1 1 1 0 01 02 03 04 05
01 0 119 178 119 Embedment ratio, d/D
6 103 164 090 (a)
20 072 144 056
60 042 125 029 5 Rough interface
100 030 120 020
Fit
025 0 144 290 150
6 109 250 083 =0
Horizontal capacity, NcH
4
20 072 221 046 =6
60 041 204 024
100 029 201 017 = 20
3
05 0 177 417 220 = 60
6 114 331 102
20 073 300 056 = 100
60 041 282 030 2
100 029 280 022
0
Table 3. Uniaxial capacity factors for skirted circular foundations 0 01 02 03 04 05
with rough soilskirt interface Embedment ratio, d/D
(b)
d=D kD=sum NcV NcH NcM
6 Rough interface Fit
0 0 605 1 067
6 985 1 119 =0
5
20 1548 1 192 =6
Moment capacity, NcM
60 2769 1 349
100 3788 1 477 4 = 20
01 0 721 178 080 = 60
6 1013 164 107 3
20 1116 144 107 = 100
60 1161 125 100 2
100 1145 120 095
025 0 871 290 101 1
6 1073 250 099
20 1108 221 088 0
60 1124 204 084 0 01 02 03 04 05
100 1104 201 081 Embedment ratio, d/D
05 0 1069 417 148 (c)
6 1122 331 121
20 1134 300 107 Fig. 10. Uniaxial bearing capacity factors for skirted circular
60 1139 282 105 foundations with rough soilskirt interface: comparison between FE
100 1116 280 103 results (discrete marker points) and fit (solid line) for (a) vertical
capacity, (b) horizontal capacity and (c) moment capacity
8 DESIGN METHOD FOR THE UNDRAINED CAPACITY OF SKIRTED FUONDATIONS
The scaling factors v, h and m are graphically rep- normalised h=h*m=m* space at a discrete level of vertical
resented in Fig. 11. The resulting bearing capacity factors for load mobilisation v 075 (h* and m* are defined later in
fully smooth skirted circular foundations are summarised in equations (15) and (16)). Fig. 13 showcases the complex
Table 4. The approximating expressions (equations (8)(13)) shape of the failure envelopes and their dependence on
provide a good fit of the numerical results, as seen in Fig. 12. foundation embedment ratio (Fig. 13(a)), soil strength
heterogeneity (Fig. 13(b)) and soilskirt interface (Fig. 13(c)).
Probe tests of u=D 05 and 05 are plotted in Figs 13(a)
Combined VHM capacity
13(c) along with their associated mechanisms of soil deforma-
Failure mechanisms. Two-dimensional slices in the HM
tion at failure. The end points of the probe tests are denoted A
plane of 3D VHM failure envelopes are shown in Fig. 13 in
for the right-hand side of the VHM failure envelope and B for
10 the left-hand side. The subscripts make reference to either
embedment ratio, soil strength heterogeneity index or soilskirt
interface. For example, A6 represents the failure point on
the right-hand side of the envelope for the skirted foundation
09
with 6. The changing shape and size of the envelopes
Scaling factor, v
0 0 605 1 067
6 985 1 119
09 20 1548 1 192
= 0, 6, 20, 60, 100
Scaling factor, m
60 2769 1 349
100 3788 1 477
08 01 0 664 148 070
6 933 136 093
20 1035 118 093
60 1082 100 086
07 100 1070 094 082
025 0 722 246 076
6 925 211 074
06 20 977 186 065
0 01 02 03 04 05 60 1001 168 060
Embedment ratio, d/D 100 985 163 058
(c) 05 0 791 365 106
6 910 289 086
Fig. 11. Scaling factors for determining the bearing capacity factors 20 933 261 075
of skirted circular foundations with smooth soilskirt interface under: 60 946 243 071
(a) pure vertical loading, (b) pure horizontal loading and (c) pure 100 929 238 069
moment loading
VULPE 9
40
Smooth interface Fit skirt (Ar) compared to the smooth skirted foundation (As),
thus increasing the size of the failure envelope.
35 =0 Embedment ratio and soil strength heterogeneity have the
30 =6 greatest influence on the shape and size of the failure
Vertical capacity, Ncv
h 1 vq 15
displacement follows an asymmetric wedge failure mechan-
ism (B20) similar to B010 and again the maximum moment is m 1 vp 16
sustained with negative horizontal load.
Figure 13(c) compares the failure envelopes as a function with p 212 and q 414. h* and m* represent the lowest
of soilskirt interface for the same level of vertical load, soil horizontal or moment normalised loads, resulted from FE
heterogeneity and embedment ratio. As expected, the size of analyses, obtained in conjunction with a known applied
the failure envelope for the fully smooth interface is lower vertical load. They are defined over the range of d=D and
than that for the rough interface. While both foundations investigated in the current study.
display similar behaviour in HM space (as illustrated by Br Values of hmax=h *, where hmax represents the maximum
and Bs), the rough skirted foundation benefits from an horizontal mobilisation as a result of HM or HM cross-
additional deep wedge mechanism around the left side of the coupling, which are required to satisfactorily reproduce
10 DESIGN METHOD FOR THE UNDRAINED CAPACITY OF SKIRTED FUONDATIONS
B050 A050
d/D = 010 Rough interface
15
d/D = 025 = 60
d/D = 050 V/Vult = 075
Probe test
B010 10 A050
B025 A025
m/m*
B025 A025
A010
05
B050
B010 A010
0
15 10 05 0 05 10 15
h/h*
(a)
10
B20
B6 A6
m/m*
B6 A0
A20 A6
B0
05
B20 A20
0
15 10 05 0 05 10 15
h/h*
(b)
10 Ar
m/m*
As
Bs As
05
Br
Bs
0
15 10 05 0 05 10 15
h/h*
(c)
Fig. 13. Effect on shape and size of VHM failure envelopes of varying: (a) foundation embedment ratio, d=D; (b) soil strength heterogeneity
index, = kD=sum; (c) soilskirt interface
10
10
m/m*
m/m*
05 05
0 0
15 10 05 0 05 10 15 15 10 05 0 05 10 15
h/h* h/h*
Fig. 14. Effect of level of vertical load on the size of the failure envelope
VULPE 11
Table 5. Fitting parameters for approximating expression for VHM envelope under vertical load mobilisation 0 V=Vult 1 for skirted circular
foundations with rough soilskirt interface
d=D
0 01 025 05 0 01 025 05
Table 6. Fitting parameters for approximating expression for VHM envelope under vertical load mobilisation 0 V=Vult 1 for skirted circular
foundations with smooth soilskirt interface
d=D
0 01 025 05 0 01 025 05
the shape of the failure envelopes, are given in Table 7 and Table 7. Values of hmax=h* for vertical load mobilisation
Table 8 for the rough and smooth interfaces, respectively. 0 V=Vult 1 for skirted circular foundations with rough soilskirt
The ability of the approximating expression to capture interface
the changing size and shape of the failure envelopes of the
skirted foundations with respect to embedment ratio, soil
strength heterogeneity and soilskirt interface is shown in
d=D 0 6 20 60 100
Fig. 15. The close fit between the FE results and the curve fits
(equation (14)) shows that the reconstruction of the failure 0 1 1 1 1 1
envelope will incur almost no error for a vertical mobilisation 01 1 1 1 107 110
v 075. The approximating expression is only slightly con- 025 111 103 101 1 1
servative for low levels of v [0, 05] for skirted foundations 05 137 117 115 113 112
of high embedment ratio d=D resting on soil with strength
heterogeneity . 0 (heterogeneous soil), but becomes increas-
ingly conservative for d=D 010 and 0 (homogeneous
soil). Nonetheless, the predictions of ultimate limit states Table 8. Values of hmax=h* for vertical load mobilisation
under combined VHM loading using equation (14) are con- 0 V=Vult 1 for skirted circular foundations with smooth soil
skirt interface
siderably less conservative than predictions resulting from the
industry recommended practice for offshore shallow foun-
dation design (DNV, 1992; API, 2000; ISO, 2003) based on
classical bearing capacity theory (as seen explicitly for ISO d=D 0 6 20 60 100
(2003) in Ukritchon et al. (1998) and Gourvenec & Barnett
(2011)). 0 1 1 1 1 1
01 1 1 109 119 122
025 102 1 1 1 1
COMPARISON OF APPROXIMATING SOLUTIONS 05 117 103 101 101 101
This study explicitly showed how embedment ratio, soil
strength heterogeneity and soilstructure interface can both
independently and in combination affect the response of a soilskirt interface (Vulpe et al., 2014) and two skirted
foundation to either uniaxial or combined loading. At the circular foundations (one with rough soilskirt interface and
same time, different foundation types (i.e. strip, solid em- the other with smooth interface) are compared where the
bedded or skirted foundations) behave distinctly under same level of embedment ratio and soil strength heterogen-
identical soil conditions and subjected to the same level of eity index are considered.
VHM loading. The importance of choosing appropriate Figure 16 summarises the moment capacity of all foun-
approximating solutions for the desired type of foundation dations with varying embedment ratio for a soil strength
geometry is further made clear in Fig. 16 and Fig. 17. The index 5. The significant scatter in the solutions is
approximating equations for the failure envelopes of a skirted evident. The combined VHM capacity is next examined for
strip foundation with rough soilskirt interface (Gourvenec solid embedded circular foundations and skirted circular
& Barnett, 2011), a solid embedded foundation with rough foundations; approximating expressions for skirted strip
12 DESIGN METHOD FOR THE UNDRAINED CAPACITY OF SKIRTED FUONDATIONS
Rough interface FE results Rough interface 15 FE results
15
d/D = 010 d/D = 025
Curve fit Curve fit
V/Vult = 075 V/Vult = 075
= 0, 6, 20, 60, 100
= 0, 6, 20, 60, 100
10 10
m/m*
m/m*
05 05
0 0
15 10 05 0 05 10 15 15 10 05 0 05 10 15
h/h*v h/h*v
m/m*
m/m*
05 05
0 0
15 10 05 0 05 10 15 15 10 05 0 05 10 15
h/h*v h/h*v
Fig. 15. Comparison of selected failure envelopes of skirted circular foundations with rough and smooth soilskirt interface predicted by FE
analyses (broken lines) and approximating expression (solid lines) (equation (14)
15
m/m*
NcM
10 05
05
=5
0 0
0 01 02 03 04 05 15 1 05 0 05 10 15
Embedment ratio, d/D h/h*v
Fig. 16. Comparison of different available solutions for determining Fig. 17. Comparison of different available approximating expressions
the pure moment capacity for determining the combined VHM capacity
VULPE 13
Table 9. Summary of proposed procedure
Step 1 Evaluate d=D, su0 and kD=sum for given foundation geometry and soil strength profile
Step 2 Calculate Fv, dcV, NcV and Vult (for skirted foundations with rough soilskirt interface) Equations (6), (3) and Table 1
Step 2 Calculate v, NcV
smooth
and Vult (for skirted foundations with fully smooth soilskirt interface) Equations (11) and (8)
Step 3 Calculate vertical mobilisation ratio v V=Vult
Step 4 Calculate h* and m* Equations (15) and (16)
Step 5 Choose exact fitting parameters and for a skirted foundation with either rough or smooth Table 5 and Table 6
soilskirt interface
Step 6 Plot normalised h*m* envelope for given v (for constant intervals of h solve for m)
Step 7 Calculate NcH and NcM (in absence of other loads or constraints) (for skirted foundations with Equations (7), (4) and (5)
rough soilskirt interface) and Table 1
Step 7 Calculate h, m, NcHsmooth smooth
and NcM (for skirted foundations with fully smooth soilskirt Equations (12), (13), (9)
interface) and (10)
Step 8 Calculate H*=Asu0 and M*=ADsu0 at given vertical load level, that is, h* and m* multiplied by
smooth smooth
NcH and NcM or NcH and NcM , respectively
Step 9 Convert h*m* envelope to dimensionless load space, that is h=(H*=Asu0) and m=(M*=ADsu0)
foundations are only available for soils with 6. The only (f) The failure envelope of the skirted circular foun-
variable in the results from Fig. 17 is the foundation type, dation cannot be approximated by a solid embedded
namely the approximation expression used, all other things circular foundation counterpart. The changes in size
being equal: foundations with embedment ratio d=D 012 and shape of the failure envelopes are greatly affected
resting on soil with 30 and subjected to a vertical load by the presence of the deformable soil plug inside
V 6 kN. The results are further compared to the industry the skirt compartment. The negative eccentricity of the
guideline solution from ISO (2003). It is noteworthy that failure envelope for low levels of embedment ratio are
all approximating expressions lead to distinct failure not captured by the approximating expression for solid
envelopes the solid embedded foundation solution does embedded circular foundations.
not capture the negative eccentricity experienced by the (g) A single algebraic expression is proposed to approxi-
skirted circular foundations. As expected, the ISO (2003) mate failure envelopes as a function of embedment
solution returns the most conservative approximation. ratio, soil strength heterogeneity index, soilskirt inter-
face and level of vertical load (the latter indirectly
through h* and m*).
CONCLUDING REMARKS
A series of small-strain FE analyses were carried out
The failure envelope methodology presented here (summar-
to explicitly investigate the effect of varying embedment
ised in Table 9) for prediction of ultimate limit states under
ratio, soil strength heterogeneity index, soilskirt interface
multi-directional loading is well suited for implementation
and level of vertical mobilisation on the combined capacity
in an automated calculation tool. Such a tool enables fast
of skirted circular foundations. The effect of all varying para-
derivation of failure envelopes of skirted circular foundations
meters was quantified through a simple algebraic approxi-
with varying embedment ratio, soil strength heterogeneity
mation expression and explained through the kinematic
index, soilskirt interface and vertical mobilisation. It avoids
mechanisms of soil deformation at failure.
the necessity to perform a large number of time-consuming
The outcomes of this study are summarised as follows.
3D FE analyses required to otherwise reproduce the failure
envelopes. The results can be used to optimise foundation
(a) The behaviour of skirted foundations differs from that geometry for a given set of design loads and shear strength
of embedded foundations. An idealisation of the skirt profile or to evaluate load or material factors.
compartment as a rigid body would lead to the over-
estimation of the moment capacities, as failure of the
soil inside the skirt compartment would not be taken ACKNOWLEDGEMENTS
into account. This work forms part of the activities of the Centre for
(b) Uniaxial bearing capacity factors are proposed through Offshore Foundation Systems (COFS), currently supported
depth factors. as a node of the Australian Research Council Centre of
(c) Scaling factors, defining the effect of soilskirt interface Excellence for Geotechnical Science and Engineering
on uniaxial bearing capacities, are formulated as a (ID CE110001009). This support is gratefully acknowledged.
function of embedment ratio and the soil strength The author is also grateful for the input provided by Mr Alex
heterogeneity index. No scaling factor for the combined Cornelius and Dr Scott Draper.
VHM failure envelope was derived owing to the
complex interaction of embedment ratio, soil strength
heterogeneity index and level of vertical mobilisation NOTATION
for either rough or fully smooth skirted foundations. A cross-sectional plan area of foundation
(d) The shape of the normalised failure envelope signifi- D foundation diameter
d skirt length
cantly varies with foundation embedment ratio and soil
dcV, dcH, dcM depth factors
strength heterogeneity index. Eu Youngs modulus
(e) The size of the normalised failure envelope varies H horizontal load
with foundation embedment ratio, soil strength hetero- Hult uniaxial horizontal capacity
geneity index, soilskirt interface and, to a lesser extent, k shear strength gradient
level of vertical load. M moment load
14 DESIGN METHOD FOR THE UNDRAINED CAPACITY OF SKIRTED FUONDATIONS
Mult uniaxial moment capacity Gourvenec, S., Acosta-Martinez, H. E. & Randolph, M. F. (2009).
smooth
NcV, NcH, NcM,NcV , bearing capacity factors Experimental study of uplift resistance of shallow skirted
smooth smooth
NcH , NcM foundations in clay under transient and sustained concentric
su shear strength loading. Gotechnique 59, No. 6, 525537, http:==dx.doi.
sum shear strength at mudline org=10.1680=geot.2007.00108.
V vertical load Hansen, J. B. (1970). A revised and extended formula for bearing
Vult uniaxial vertical capacity capacity, Bulletin 28, pp. 511. Copenhagen, Denmark: Danish
z depth Geotechnical Institute.
effective unit weight Houlsby, G. T. & Wroth, C. P. (1983). Calculation of stresses
v, h, m scaling factors on shallow penetrometers and footings. Proceedings of
soil heterogeneity index IUTAM=IUGG seabed mechanics, Newcastle, pp. 107112.
Poisson ratio House, A. R. & Randolph, M. F. (2001). Installation and pull-out
capacity of stiffened suction caissons in cohesive sediments.
Proceedings of the 11th international offshore and polar
REFERENCES engineering conference, Stavanger, Norway, vol. 2, pp. 574580.
Andersen, K. H. & Jostad, H. P. (2002). Shear strength along ISO (International Standardisation Organisation) (2003).
outside wall of suction anchors in clay. Proceedings of the ISO 19901-4: Petroleum and natural gas industries specific
31st offshore technology conference, Houston, TX, paper OTC requirements for offshore structures Part 4: Geotechnical
10824. and foundation design considerations, 1st edn. Geneva,
API (American Petroleum Institute) (2000). API RP 2A: Switzerland: International Standardisation Organisation.
Recommended practice for planning, designing and constructing Jostad, H. P. & Andersen, K. H. (2006). Potential benefits of using
offshore platforms, 21st edn. Washington, DC, USA: American skirted foundations for jackup platforms. Proceedings of the
Petroleum Institute. offshore technology conference, Houston, TX, paper, OTC
Barari, A. & Ibsen, L. B. (2012). Undrained response of bucket 18016.
foundations to moment loading. Appl. Ocean Res. 36, 1221. Kellezi, L., Hofstede, H. W. L. & Hansen, P. B.
Bienen, B., Gaudin, C., Cassidy, M. J., Rausch, L., Purwana, O. A. (2005). Jack-up footing penetration and fixity analyses.
& Krisdani, H. (2012). Numerical modelling of a hybrid skirted Proceedings of the 1st international symposium frontiers
foundation under combined loading. Comput. Geotech 45, in offshore geomechanics (ISFOG), Perth, Australia,
127139 pp. 559565.
Bransby, M. F. & Randolph, M. F. (1998). Combined loading Kellezi, L., Kudsk, G. & Hofstede, H. W. L. (2008). Skirted footings
of skirted foundations. Gotechnique 48, No. 5, 637655, http:== capacity for combined loads and layered soil conditions. In
dx.doi.org=10.1680=geot.1998.48.5.637. Foundations: Proceedings of the 2nd British Geotechnical
Bransby, M. F. & Randolph, M. F. (1999). The effect of embedment Association international conference on foundations (eds
depth on the undrained response of skirted foundations to M. J. Brown, M. F. Bransby, A. J. Brennan and J. Knappett),
combined loading. Soils Found. 39, No. 4, 1933. vol. 1, pp. 923935. Dundee, UK: HIS Press.
Bransby, M. F. & Yun, G. J. (2009). The undrained capacity of Kourkoulis, R. S., Lekkakis, P. C., Gelagoti, F. M. & Kaynia, A. M.
skirted strip foundations under combined loading. Gotechnique (2014). Suction caisson foundations for offshore wind turbines
59, No. 2, 115125, http:==dx.doi.org=10.1680=geot.2007. subjected to wave and earthquake loading: effect of soil
00098. foundation interface. Gotechnique 64, No. 3, 171185, http:==
Britto, A. M. & Kusakabe, O. (1982). Stability of unsupported dx.doi.org=10.1680=geot.12.P.179.
axisymmetric excavations in soft clay. Gotechnique 32, No. 3, Larsen, K. A., Ibsen, L. B. & Barari, A. (2013). Modified
261270, http:==dx.doi.org=10.1680=geot.1982.32.3.261. expression for the failure criterion of bucket foundations
Butterfield, R., Houlsby, G. T. & Gottardi, G. (1997). Standardised subjected to combined loading. Can. Geotech. J. 50,
sign conventions and notation for generally loaded foundations. No. 12, 12501259.
Gotechnique 47, No. 5, 10511052, http:==dx.doi.org=10. Mana, D. K. S., Gourvenec, S., Randolph, M. F. & Hossain, M. S.
1680=geot.1997.47.5.1051. (2012). Failure mechanisms of skirted foundations in uplift
Chen, W. & Randolph, M. F. (2007). External radial stress changes and compression. Int. J. Phys. Modelling Geotech. 12, No. 2,
and axial capacity for suction caissons in soft clay. Gotechnique 4762.
57, No. 6, 499511, http:==dx.doi.org=10.1680=geot.2007.57. Mana, D. K. S., Gourvenec, S. & Randolph, M. F. (2013).
6.499. Experimental investigation of reverse end bearing of offshore
Cox, A. D., Eason, G. & Hopkins, H. G. (1961). Axially symmetric shallow foundations. Can. Geotech. J. 50, No. 10, 10221033.
plastic deformation in soils. Proc. R. Soc. London, Ser. A 254, Martin, C. M. (2003). New software for rigorous bearing
No. 1036, 145). capacity calculations. Proceedings of the British Geotechnical
Dassault Systmes (2012). Abaqus analysis users manual. Association international conference on foundations, Dundee,
Providence, RI, USA: Simulia Corp. pp. 581592.
DNV (Det Norske Veritas) (1992). Foundations, classifications notes Martin, C. M. & Randolph, M. F. (2001). Applications of the lower
no. 30.4. Oslo, Norway: Det Norske Veritas. and upper bound theorems of plasticity to collapse of circular
Gourvenec, S. (2007). Failure envelopes for offshore shallow foun- foundations. Proceedings of the 10th international conference
dations under general loading. Gotechnique 57, No. 9, 715728, of the International Association of Computer Methods
http:==dx.doi.org=10.1680=geot.2007.57.9.715. and Advances in Geomechanics (IACMAG), Tucson, vol. 2,
Gourvenec, S. (2008). Effect of embedment on the undrained pp. 14171428.
capacity of shallow foundations under general loading. Meyerhof, G. G. (1953). The bearing capacity of foundations
Gotechnique 58, No. 3, 177185, http:==dx.doi.org=10. under eccentric and inclined loads. Proceedings of the
1680=geot.2008.58.3.177. 3rd international conference on soil mechanics and foundation
Gourvenec, S. & Barnett, S. (2011). Undrained failure envelope engineering, Zurich, vol. 1, pp. 440445.
for skirted foundations under general loading. Gotechnique 61, Puech, A., Iorio, J.-P., Garnier, J. & Foray, P. (1993). Experimental
No. 3, 263270, http:==dx.doi.org=10.1680=geot.9.T.027. study of suction effects under mudmat type foundations.
Gourvenec, S. & Mana, D. K. S. (2011). Undrained vertical Proceedings of Canadian conference on marine geotechnical
bearing capacity factors for shallow foundations. Gotechnique engineering, St Johns, Newfoundland, vol. 3, pp. 10621080.
Lett. 1, OctoberDecember, 101108, http:==dx.doi. Randolph, M. F. & Puzrin, A. M. (2003). Upper bound limit
org=10.1680=geolett.11.00026. analysis of circular foundations on clay under general loading.
Gourvenec, S. & Randolph, M. F. (2003). Effect of strength Gotechnique 53, No. 9, 785796, http:==dx.doi.org=
non-homogeneity on the shape and failure envelopes for com- 10.1680=geot.2003.53.9.785.
bined loading of strip and circular foundations on clay. Skau, K. S. & Jostad, H. P. (2014). Application of the
Gotechnique 53, No. 6, 575586, http:==dx.doi.org=10. NGI-procedure for design of bucket foundations for
1680=geot.2003.53.6.575. offshore wind farms. Proceedings of the 24th international
VULPE 15
ocean and polar engineering conference, Busan, Korea, Ukritchon, B., Whittle, A. J. & Sloan, S. W. (1998). Undrained
pp. 189198. limit analysis for combined loading of strip footings on clay.
Skempton, A. W. (1951). The bearing capacity of clays. Proceedings J. Geotech. Geoenviron. Engng, ASCE 124, No. 3, 265276.
of the building and research congress, London, vol. 1, Vulpe, C., Bienen, B. & Gaudin, C. (2013). Predicting the undrained
pp. 180189. capacity of skirted spudcans under combined loading. Ocean
Supachawarote, C., Randolph, M. F. & Gourvenec, S. (2005). The Engng 74, 178188.
effect of crack formation on the inclined pullout capacity of Vulpe, C., Gourvenec, S. & Power, M. (2014). A generalised failure
suction caissons. Proceedings of the International Association of envelope for undrained capacity of circular shallow foundations
Computer Methods and Advances in Geomechanics conference, under general loading. Gotechnique Lett. 4, No. 3, 187196.
Turin, Italy, pp. 577584. Rotterdam, the Netherlands: Watson, P. G., Randolph, M. F. & Bransby, M. F. (2000). Combined
Balkema. lateral and vertical loading of caisson foundations. Proceedings
Taiebat, H. A. & Carter, J. P. (2000). Numerical studies of of the annual offshore technology conference, Houston, TX,
the bearing capacity of shallow foundations on cohesive paper OTC 12195.
soil subjected to combined loading. Gotechnique 50, Yun, G. & Bransby, M. F. (2007). The horizontal-moment capacity
No. 4, 409418, http:==dx.doi.org=10.1680=geot.2000.50.4. of embedded foundations in undrained soil. Can. Geotech. J. 44,
409. No. 4, 409427.