ABSTRACT:

Offshore wind turbines are considered as a reliable source of renewable energy in the UK but there are no long term
observations of the performance of these relatively novel structures. The structures are dynamically sensitive due to the very
nature of the shape and mass distribution (slender and top heavy). Monitoring of a limited number of offshore wind turbines has
indicated a gradual departure of the overall system dynamics from the design requirements. The aim of this paper is to carry out
rigorous analytical studies to establish the natural (linear) frequencies and resonant (nonlinear) frequencies of a pile-soil system.
Numerical continuation is used to identify nonlinear frequency response functions of the resulting Duffings like nonlinear
oscillator. The back bone curves of these plots enable an empirical formulae for the resonant system frequency with input
amplitude. The presence of subharmonic resonances is explored.
KEY WORDS: Piles, nonlinear dynamics, P-y curves, subharmonic resonances, Duffings oscillator
1 INTRODUCTION
Offshore wind turbines are currently considered as a reliable
source of renewable energy in the UK. These structures are
dynamically sensitive as their natural frequency is very close
to the various forcing frequencies imposed upon them either
by the environmental loads (strong wind and wave loading) or
by the on-board machinery including the rotor and its
associated effects. Figure 1 shows a schematic diagram of an
offshore wind turbine with a typical wind distribution along
the length of the tower for two relative positions of the blades
with respect to the tower. The tower, above the water,
experiences two types of loads: (a) the bottom part of the
tower, unobstructed by the spinning turbine blade experiences
a nearly constant value of the wind loading; (b) the top part of
the tower, which is periodically obstructed by the spinning of
the blades is subjected to a cyclic loading often called the
blade passing effect (2P/3P depending on 2 blades or 3 blades
respectively) or blade shadowing effect or wind shielding
effect in the literature.

Figure 1: Figure showing the effect of blade shadowing

Figure 2 shows the main frequencies for a three-bladed 3MW
Vestas V90 Wind turbine with an operational interval of 8.6 to
18.4rpm: the rotor frequency (often termed as 1P) lies in the
range 0.14-0.3Hz and the corresponding blade passing
frequency for a three-bladed turbine lies in the range 0.42-
0.9Hz. The figure also shows a typical frequency distribution
for wind and wave loading. The peak frequency of offshore
waves is about 0.1Hz.


Figure 2: Power Spectral density plotted against the
various forcing frequency
It is clear from Figure 2 that the designer has to select a
system frequency (the global frequency of the overall wind
turbine-foundation system) which lies outside these in order to
avoid system resonance type effects. The usual choice would
lie between turbine and blade passing frequencies (so-called
soft-stiff, option 2 in Figure 2). There are two challenges:
1. The foundation stiffness must be estimated very
accurately from the available soil data.
2. The potential for change in foundation stiffness with
time as a result of the cyclic loading must be
understood so that the risks of the system frequency
coinciding with a loading frequency can be avoided.
This makes the design of foundations extremely challenging.
Figure 3 shows the schematic of a offshore wind turbine
supported on a single tubular section inserted deep into the
The dynamics of monopile-supported Wind Turbines in nonlinear soil
N.A.Alexander
1
, S.Bhattacharya
2


1
Department of Civil Engineering, University of Bristol, UK
2
Department of Civil Engineering, University of Bristol, UK

Proceedings of the 8th International Conference on Structural Dynamics, EURODYN 2011
Leuven, Belgium, 4-6 July 2011
G. De Roeck, G. Degrande, G. Lombaert, G. M uller (eds.)
ISBN 978-90-760-1931-4
3416
ground known as monopile. The total environmental lateral
load acting on the offshore wind turbine in Figure 3 can be
modeled simplistically as an instantaneous static horizontal
load, P acting at a distance h above the foundation level.
Thus P represents the resultant lateral load on the tower that
must be resisted by the monopile foundation. Figure 3 also
shows an equivalent force model for the foundation where the
lateral load P on the tower is replaced by a force P and a
moment Ph at the pile head.



Figure 3: Schematic diagram of the monopile-supported
wind turbine

Experimental studies of the overall system using a 1:100 scale
model has been carried out by Lombardi (2010), Bhattacharya
et al (2011). In these experimental studies, a scaled model is
constructed and the environmental dynamic loads which acts
on a prototype offshore wind turbine is imposed using an
electro-dynamic actuator fixed to the laboratory strong wall at
BLADE (Bristol Laboratory for Advanced Dynamics
Engineering) and connected to the model wind turbine tower.
The force applied to the wind turbine could be constantly
monitored by a force sensor. The blades were rotated by a DC
electric motor to model the 1P loading and also to give some
aerodynamic damping to the system. A typical test consists of
the application of the cyclic loading for a particular time
interval (or certain number of cycles) and then measuring the
frequency and damping of the system by a free vibration test.
In the free vibration test (also known as a snap back test in
the literature), the actuator was disconnected from the tower
and the tower was given a small amplitude vibration and the
acceleration of the system recorded. The test results were
analyzed using mechanics-based non-dimensional groups
derived in Bhattacharya et al (2011). The test results showed
that under certain loading conditions, the natural frequency of
the system changes as a result of the effects of cyclic loading
and. These critical conditions relate to: (a) the strain level in
the soil which is dependent of the lateral load (P) acting on the
foundation, shear modulus of the soil and the dimensions of
the foundation resisting the load; (b) the relative position of
the system frequency in comparison to the forcing frequency;
(c) Number of cycles.
Aim and scope of the paper:
It appears from the limited experimental results that dynamics
is an important design consideration. To make more sense of
the experimental data, a rigorous analytical study has been
carried out to determine the following:
a) The small strain (linear) natural frequency of the system
b) The reduction of this resonant frequency with increases
in forcing amplitude. (i.e. the nonlinear resonance
frequency back bone curves)
c) The sub-harmonic resonant frequency backbone curves.

Thus we identify all the frequencies that the system may
experience resonances for both small and large soil strain
levels.

2 ANALYTICAL MODEL

Consider the system described in Figure 1. The pile is
surrounded by a near field soil stack of layers. A
continuum of layers surround the pile. Each layer contains a
nonlinear spring, nonlinear dashpot and lumped masses. The
effective width of the near-field layer is
w
L . The system is
sub-structured such that only the near field soil, pile and
superstructure mass are considered. The displacement of the
pile is ( ) , y z t .

Figure 1: Analytical model

2.1 Using P-y backbone curves
For a range of different soils many different P-y curves
proposed by researchers [1-6]. The fidelity of these forms of
curves to experimental data has been questioned by [6, 7].
Results suggest that there is some scatter about these
nonlinear trends. This scatter is due to lack of heterogeneity
and complex hysteretic nonlinear behavior of real soils. An
example of this scatter is given in [8, 9]; where full scale
dynamic test data for a pile driven in medium-dense sand is
obtained. Dynamic test results show an approximately cubic
p-y curve (where p is soil spring load per unit area ) complete
with a falling part, (c.f. fig 12 in [8]) .

Displacement
( ) , y z t
L
z

h
m
0
y

Nonlinear
Spring
Nonlinear
damper
Lumped mass
y

Soil
layer
h
M Ph =
Proceedings of the 8th International Conference on Structural Dynamics, EURODYN 2011 3417
Note that P-y curves for pile-soil springs are normally
obtained by applying a load at the head of the pile and
recording the displacements of the pile. This is the case for
experimental study [8, 9]. More recent work, [10], has
suggested that the p-y curves can be loading rate dependant.
Thus [10] suggest p-y curves based on static or slow cyclic
loading may underestimate the backbone observed under
faster cyclic loading. This strain rate hardening in clay soils is
clear; in sands this effect seems less marked. The basic shape
of these cyclic p-y curves, in [10], is still fairly similar to the
static p-y curves with a scaling (stretching) of the p axis.

However, for all their limitations P-y curves provide a
pragmatic and useful engineering simplification and are used
widely in practice. In this paper the form of [6, 11] is taken.
This is shown in equation (1); where P is the soil-spring
reaction at depth z,
ub
P is the ultimate bearing capacity at the
base of the pile i.e. at depth L, A is 0.9 a reduction factor for
cyclic behavior,
0
k is the initial modulus of subgrade reaction
[5], y is the horizontal pile deflection at depth z. This follows
the functional form presented in work of [12-15]. This linear
assumption differs from the American Petroleum Institute
(API) [11] recommendation that assumes a quadratic variation
of P with depth for shallow depths and a linear variation
of P with depth for deep depths. [15] suggests that the API
quadratic variation that was based on a wedge failure
mechanism at shallow depths is unduly optimistic compared
with other proposed models. Experimental results presented in
[16] (c.f. figs 12 and 14 ) show a near linearly increase in
u
P
with depth. Nevertheless, in this paper, a linear variation with
depth is assumed mainly because it is computationally
simplifying.
( ) ( )
0
, tanh
z
ub L
ub
k L
P x z AP y
AP
| |
=
|
\ .
(1)

For dynamic p-y curves to be incorporated, as [10] suggests, a
the factor A should be scaled. A Taylor series expansion
suggests that the tanh function can be expressed as a cubic
with higher terms neglected, for a small range of y. However,
just expanding and neglecting higher order terms does not
often produce the best fit over the problem domain
p
y y s .
The optimal values of linear and nonlinear spring constants
1 w
k and
2 w
k should be sought, in a least square sense. Thus,
the optimal spring coefficients are given by (3) and (4) in
terms of
0
k , L,
p
y and
ub
P . Note that for
p
y y > the cubic
function falls away from API backbone curve. This can be an
advantage as it models degradation of stiffness at large cyclic
strains as seen in [8].

In summary, the cubic P-y curve proposed in this paper
enforces quartic nonlinear strain energy of the soil.


( ) ( )
3 5
1 2
z
w w L
P k y k y O y ~ + ,
p
y y s (2)

( ) ( ) 3 5
3 105
1
4
0
5
4
7
0
p p
p p
y y
L L
w z z
y y
k Py dy Py dy =
} }
(3)

( ) ( ) 5 7
3 105 175
2
4 4
0 0
p p
p p
y y
L L
w z z
y y
k Py dy Py dy =
} }
(4)

2.2 Potential energy of system
The potential Energy U of this system is composed of two
terms (i) the flexural strain energy of the pile group and (ii)
the soil spring stiffness energy. The work done by the far field
soil reactions Q is neglected here as it does not contribute to
the subsequent equations of motion. Hysteretic behavior of
springs is not modeled directly as this greatly increases the
analytical complexity. However, it is not ignored. The
hysteretic behavior of springs is summed up per cycle and
included as an increased nonlinear damping term proposed
later. This approach follows work of Voigt and Maxwell
described in [17]. This approach is used to determine the
variation of soil damping coefficient with strain levels.


( ) ( )
( ) ( )
2 1
2
0
2 4 1 1
1 2 2 4
0
d 0 0
d
L
L
z z
w w L L
U EIy z Py Phy
k y k y z
'' ' =
+
}
}
(5)

If a classical Rayleigh-Ritz spatial-temporal series is adopted,
i.e.
i i
y w| = E then an n degree of freedom nonlinear system is
obtained. This is equivalent to a high order (i.e. 2n) nonlinear
ODE. Unfortunately, this is not suitable for an initial
investigation into the problem. In this paper, a first order
approximation is assumed ( ) ( ) y Lw t | = that result in a
nonlinear, single degree of freedom system. This introduces
one unknown degree of freedom, w , where wL is the
displacement at the top of the pile. A non-dimensional
variable, z L = , is introduced; it is an ordinate along the
pile. Hence primes and double primed variables from now on
denote first and second derivative with respect to .

( )
2 4 1 1
1 2 2 2
EI
L
U K w K w Pw = (6)

4 6
1 2
1 2
,
w w
k L k L
EI EI
q q = = (7)
1
2 2
1 1
0
d K | q | '' = +
}
,
1
4
2 2
0
d K q | =
}
(8)
2.3 Kinetic energy of system
The translational, horizontal, kinetic energy T of the system is
defined in (10). The rotational kinetic energy of pile and
superstructure mass are neglected in this paper. Parameter
h
m
is the lumped mass at pile head; i.e. the mass of any structure
attached to the top of the pile. m is the mass per unit depth of
the pile and soil layer; where
p
m is the mass per unit depth of
the pile and
s
density of the near field soil layer. The out
of plain breadth of the soil layer is assumed to be
b
D . The
soil mass for a layer z o is lumped half on the pile and half at
the near/far field boundary. The kinetic energy can be re-
expressed, in quadratic form, in terms of degrees of freedom
w by equations (10)

Proceedings of the 8th International Conference on Structural Dynamics, EURODYN 2011 3418

1
2 p s w b
m m LD = + (9)
( )
2
2 3 2 1 1 1
2 2 2
0
0 d =
L
h
T m y m y z mL Mw = +
}
(10)
( )
1
2
2
0
0 d M o| | = +
}
(11)

h
m
mL
o = (12)

2.4 Rayleigh dissipative function of system
The damping coefficient of the soil is assumed quadratic in
nature, (13); this follows the experimental evidence of [7, 18]
that suggest the ratio of critical damping is dependent of
cyclic soil shear strain amplitude. The strain in the soil spring
is
w
y L c = . The system Rayleigh dissipative function, R, is
given by equation (14); frequency s is introduced to aid later
simplifications.


2
1 2
c c c c = + (13)
( )
( )
( )
2
2 3 2 2 1 1
1 2 1 2 2 2
0 w
L
y
L
R c c y dz mL s C C w w

= + = +
}
(14)

1 1
2 4 1 2
1 2 2
0 0
d , d
w
c c
C C
ms ms
| |

= =
} }
(15)

2
4
EI
s
mL
= (16)
2.5 Equation of motion for system
Thus the Euler-Lagrange-Rayleigh equation of motion (17)
can be obtained from (6) (10) and (14). This is re-expressed
by the introduction of frequency parameter
1
e equation (18).
Also non-dimensional parameters | ,
1
,
2
are introduced,
see equations (19)

( )
2 3
1 2
2 w w w w w P | + + + = (17)

1
1
K
s
M
e = (18)

2
1
K
K
| = ,
1
1
1
2
C
MK
=
2
2
1
2
C
MK
= (19)

This equation of motion (17) has been further simplified by
introducing a time-scale
1
t t e = . Note, subsequently, dots
above letters denote differentials with respect to t i.e.
w dw dt = and
2 2
w d w dt =

One final further scaling simplifies this equation. The
generalised ordinate
1 2
w u |

= . This results in equation (20),
that is a Duffings [19-22] oscillator but with nonlinear
damping.

( )
2 3
1
2 1
n
u u u u u F + + + = (20)

2
1
n

|
= ,
1 2
F P | = (21)

3 SOLUTIONS OF NONLINEAR SYSTEM
3.1 Linear natural frequency of system
To determine the linear natural frequency of the system
Equation (18) must be used; however, this requires a
knowledge of the shape function | . So what function should
be used for | ? For the linear system we can determine the
first mode eigenvector and use this for | . For the solution of
the nonlinear system we would ideally like | to be closer to
the fundamental nonlinear mode shape, if such a thing can be
defined. In this case using the linear mode would represent a
projection of the nonlinear system onto a truncated linear
modal basis. It would give a very good approximation at low
amplitudes. It would be useful to observe how the soil/pile
linear stiffness ratio,
1
q , influences the linear fundamental
natural mode| .
-0.5 0 0.5 1 1.5
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
| (deflection)

(
d
e
p
t
h
)


-80 -60 -40 -20 0 20
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
|'' (moment/EI)
-500 0 500 1000
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
|''' (shear/EI)
q
1
=10
2
q
1
=10
4
q
1
=10
6

Figure 2: First linear mode for pile with cap

By adopting the method described in [23] we can determine
the fundamental mode shapes for a pile with a free toe
(floating toe) and fixed head. Figure 2 shows the variation in
mode shape with parameter
1
q which is the soil to pile
stiffness ratio. From these figures the influence of
1
q on the
deflected shape is significant. However, for a given
1
q , the
reduction in soil-springs stiffness that is due to nonlinear
behavior of the soil changes the deflected shape a little. This
is because only a change in the order of magnitude of
1
q
effects the first mode shapes significantly. Thus, there is
evidence to suggest, in this case, that the proposed single
shape function model is a reasonable first approximation to
the nonlinear natural mode. In addition, this is evidence that
suggests the use of a single degree of freedom reduced order
model is reasonable and valid. As a further validation the
moment and shear force diagrams are very similar,
qualitatively, to those obtained experimentally [7-9, 18] and
are consistent with other theoretical work [24].

From the same eigenproblem is it straightforward to
determine the variation of first linear natural frequency
1
e
Proceedings of the 8th International Conference on Structural Dynamics, EURODYN 2011 3419
with parameter
1
q (soil/pile stiffness ratio) and o (mass at
head of pile to soil-pile mass ratio); this is displayed in Figure
2. (c.f. [23])


10
0
10
1
10
2
10
3
10
4
10
5
10
6
10
-1
10
0
10
1
10
2
10
3
0
0.5
1
1.5
2
q
1
e
1

/

s
0
0.5
1
1.5 2
o value for pile without cap
o value for pile
with cap

Figure 3: linear fundamental frequency
1
e vs. Soil/pile
stiffness ratio
1
q
3.2 Nonlinear frequency response functions
In order to obtain a nonlinear resonance curve the system is
driven by harmonic forcing, hence the equation of motion is
(22). F is the amplitude of the forcing in non-dimensional
equation. The non-dimensional forcing frequency ratio e
equals the ratio
1 f
e e where
f
e would be the dimensional
forcing frequency.

( ) ( )
2 3
1
2 1 sin
n
u u u u u F et + + + = (22)

The parameter | is dependent on the shape functions | that
is influenced by the boundary conditions (constraints) of the
pile. However it appears, at first glance, that these play no part
in equation (22)

In order to solve equation (22) damping parameters
1
and
n
must be specified. A significant amount of
experimental work [25-28] and others has been performed on
soils in order to determine the relationship between damping
ratio and cyclic shear strain amplitude. Most soils have a
damping ratio of 0.01 to 0.02 at low strain levels (of the order
of 10
-5
) this grows approximately quadratically to about 0.15
to 0.25 at larger strain levels (of the order of 10
-2
). Note that
the normalized cubic stiffness term of (22) is bounded
physically by 1 u s . Exceeding this limit is considered failure
of the soil. Thus, it is assumed that the maximum damping of
the soil is achieved at this maximum displacement; hence
( )
max 1
1
n
= .

The nonlinear frequency response function can be determined
many ways; here we use a full numerical continuation
algorithm [29]. This is more accurate than the harmonic
balance approach in [23] and can also capture the
subharmonic resonances missed by previously. Figure 4 is an
example of these frequency response functions for a range of
difference forcing amplitude F. The damping parameters
employed here were
1
0.01 = and
max
0.1 = (i.e. the
nonlinear damping varied between 1% and 10% of critical
damping)

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Forcing Frequency Ratio e
R
e
s
p
o
n
s
e

A
m
p
l
i
t
u
d
e
,

m
a
x
|
u
|
Subharmonic
Resonance
Primary Nonlinear
Resonance Backbone
0.075
0.025
0.015
F=0.005 0.005


Figure 4: Nonlinear frequency response function including
primary and subharmonic resonances.
In this figure we observe the classical soften primary
resonance peak that bends over to the left. This indicates a
reduction in resonant frequency with increasing forcing
amplitude. One feature previously missing in [23] is the
presence of a series of branch points at very high amplitude
response on the top parts of this resonance feature. These
indicate the presence of much more complex dynamics than
was observed using a harmonic balance approach. However,
as the amplitude is nearly one the system is very nearly at
failure and well beyond the accurate range of the cubic fit of
the tanh function for the P-y soil spring stiffness relationship.
Hence these more complex feature are extremely unlikely to
have an physical interpretation here.

Figure 4 also shows the presence of the subharmonic
resonances; around 0.326 e ~ . This subharmonic resonance is
only observed when the forcing amplitude 0.03 F > which is
a very large forcing amplitude; (if one considers the primary
resonance). Note that this subharmonic grows with increasing
forcing amplitude but is significantly smaller than the primary
resonance. It seems unlikely that this feature will effect, to
any great extend, the response of a pile system under working
loads. Thus the primary resonance is all than needs to be
considered.

Proceedings of the 8th International Conference on Structural Dynamics, EURODYN 2011 3420
3.3 Formula for nonlinear resonant frequency
Figure 5 is a plot of the primary nonlinear resonant
frequencies
r
e vs. forcing amplitude F for various
n
. Note
that
1
shall be, in this paper, universally assumed 0.01.
Empirical equation (23) is a good least square fit for the
nonlinear resonant frequencies (primary) obtained from the
frequency response functions.


1
3
2 6
1 7.163 0.1433
0.6578 4.808 10 0.6
r
n
n
n
F F
F
F
e

e

~ +
| |
>
|
\ .
(23)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.6
0.7
0.8
0.9
1
1.1
R
2
=0.99696
F
e
r

/
e
1
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0
0.2
0.4
0.6
0.8
1
R
2
=0.99713
F
|
|
u
|
|

25%
25%
16% 19% 22%

max
= 10%
13%
max
= 10%

Figure 5: Variation of nonlinear resonant frequency
r
e with
forcing amplitude F.

5 CONCLUSIONS

The paper presents:
(i) a chart (Figure 3) for the variation of linear natural
frequency of a pile with pile/soil stiffness ratio
1
q and
mass ratio o
(ii) an equation (23) that relates the nonlinear resonant
frequency vs. non-dimensional forcing amplitude F
with the dimension forcing amplitude P by
F P | = where | is the ratio of nonlinear to linear
soil stiffnesses, see equations (19) and (8).

With these results it is possible to estimate the nonlinear
resonant frequency of a pile in nonlinear soil.

The presence of nonlinear subharmonic resonances at about
1/3
rd
the natural frequency was explored. While these
nonlinear features exist it is clear that they occur only at very
high forcing amplitude levels that are likely to fail the
structure at the primary resonance. Additionally the
amplification factor at these subharmonic resonances is far
smaller than those at the primary resonance. Hence it appears
that these subharmonic resonances are unlikely to play a
major role in the dynamics of real pile systems.
REFERENCES

[1] R. L. Kondner, "Hyperbolic stress-strain response:
Cohesive soils," J. Soil Mech. Found. Dicv, vol. 89,
pp. 115-144, 1963.
[2] "Det Norske Veritas, Rules for the design,
construction, and inspection of offshore structures,
Appendix F; Foundations, (Hovik, Norway)," 1980.
[3] R. F. Scott, "Analysis of centrifuge pile tests;
Simulation of pile driving," American Petroleum
Institute, Washington DC, USA1980.
[4] H. Matlock, "Correlations for design of laterally
loaded piles in soft clay," presented at the Proc. 2nd
Annual Offshore Technol. Conf., Houston, USA,
1970.
[5] L. C. Reese, et al., "Analysis of laterally loaded piles
in sand," presented at the Proc. 6nd Annual Offshore
Technol. Conf., Houston, USA, 1974.
[6] J. M. Murchison and M. W. O'Neill, "evaluation of
p-y relationships in cohesionless soils," San
Francisco, CA, USA, 1984, pp. 174-191.
[7] W. Finn, "A study of piles during earthquakes: Issues
of design and analysis," bulletin of earthquake
engineering, pp. 141-234, 2005.
[8] j. m. ting, et al., "centrifuge static and dynamic
lateral pile behaviour," Canadian Geotechnical
Journal, vol. 24, pp. 198-207, 1987.
[9] J. M. Ting, "full-scale cyclic dynamic lateral pile
responses," Journal of Geotechnical Engineering,
vol. 113, pp. 30-45, 1987.
[10] M. H. El Naggar and K. J. Bentley, "Dynamic
analysis for laterally loaded piles and dynamic p-y
curves," Canadian Geotechnical Journal, vol. 37, pp.
1166-1183, Dec 2000.
[11] API, "Recommended Practice for Planning,
Designing and Constructing Fixed Offshore
Platforms - Working Stress Design - Includes
Supplement 2 " American Petroleum Institute2000
[12] J. Brinch Hansen, "The ultimate resistance of rigid
piles against transversal forces," Danish
Geotechnical Institute, Copenhagen, Denmark, vol.
12, pp. 5-9, 1961.
[13] W. G. K. Fleming, et al., Piling Engineering: Surrey
University Press, London, 1992.
[14] L. Zhang, et al., "Ultimate lateral resistance to piles
in cohesionless soils," Journal of Geotechnical and
Geoenvironmental Engineering, vol. 131, pp. 78-83,
2005.
[15] M. Randolph, "The challenges of offshore
geotechnical engineering," Ground Engineering, vol.
38, pp. 32-33, 2005.
[16] B. T. Kim, et al., "Experimental load-transfer curves
of laterally loaded piles in Nak-Dong River sand,"
Journal of Geotechnical and Geoenvironmental
Engineering, vol. 130, pp. 416-425, 2004.
[17] Dynamic analysis and earthquake resistant design,
Strong motion and Dynamic Proerties, Japanese
Society of Civil Engineers vol. 1: A.A. Balkema,
rotterdam, Brookfield, 1997.
Proceedings of the 8th International Conference on Structural Dynamics, EURODYN 2011 3421
[18] L. Verdure, et al., "Lateral cyclic loading of single
piles in sand," International Journal of Physical
Modelling in Geotechnics, vol. 3, pp. 17-28, 2003.
[19] P. J. Holmes and D. A. Rand, "bifurcations of
duffings equation - application of catastrophe
theory," Journal of Sound and Vibration, vol. 44, pp.
237-253, 1976.
[20] Y. Ueda, "randomly transitional phenomena in the
system governed by duffings equation," Journal of
Statistical Physics, vol. 20, pp. 181-196, 1979.
[21] C. Pezeshki and E. H. Dowell, "on chaos and fractal
behavior in a generalized duffings system," Physica
D, vol. 32, pp. 194-209, Sep 1988.
[22] F. N. H. Robinson, "experimental-observation of the
large-amplitude solutions of duffings and related
equations," Ima Journal of Applied Mathematics,
vol. 42, pp. 177-201, 1989.
[23] N. A. Alexander, "Estimating the nonlinear resonant
frequency of a single pile in nonlinear soil," Journal
of Sound and Vibration, vol. 329, pp. 1137-1153,
2010.
[24] L. C. Reese and H. Matlock, "Non-dimensional
solutions for laterally loaded piles with soil modulus
assumed proportional tp depth," in Proc. 8th Texas
Conference on Soil Mechanics and Foundation
Engineering, Austin Texas., 1956, pp. 1-41.
[25] T. Koskusho, "Dynamic deformation properties of
soil and nonlinear seismic response.," Central
Research Institute of Electric Power Industry (Japan)
301, 1982.
[26] Nishi, et al., "Dynamic properties of soil-structure
system for JPDR," Central Research Institute of
Electrip Power Industry (Japan) 383002, 1983.
[27] B. O. Hardin and V. P. Drnevich, "Shear modulus
and damping in soils: measurement and parameter
effects," American Society of Civil Engineers,
Journal of the Soil Mechanics and Foundations
Division, vol. 98, pp. 603-624, 1972.
[28] M. L. Silver and H. B. Seed, "Deformation
characteristics of sands under cyclic loading,"
Journal of the Soil Mechanics and Foundations
Division, vol. 97, pp. 1081-98, 1971.
[29] N. A. Alexander, "Computational algorithms for the
global stability analysis of driven oscillators," Ph.D,
Civil Engineering, University College, London.,
London, 1990.


Proceedings of the 8th International Conference on Structural Dynamics, EURODYN 2011 3422