Stability Analysis of Tidal Current Turbine

Foundation
Annual Report 2014 / 2015

Xiaojun Tang - PhD Student in Civil Engineering

International Academy for the Marine Economy and Technology
[email protected]

Supervisers:
Dr. Juan Wang
Dr. Xia li
Dr. Fangfang Zhu
Contents

1 Basis 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Aims,objectives and outputs . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.1 Reflection on existing projects . . . . . . . . . . . . . . . . . . 4
1.4.2 Loading calculation . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.3 Finite-element simulation . . . . . . . . . . . . . . . . . . . . 4

2 Literature Review 6
2.1 Tidal Current Turbine . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.1 Tidal Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.2 Offshore Tidal Turbine . . . . . . . . . . . . . . . . . . . . . . 8
2.1.3 Supporting Structure . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.4 Foundations and Anchoring System . . . . . . . . . . . . . . . 13
2.2 Stability analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.1 Slip-line method . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.2 Limit equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.3 Displacement finite-element analysis . . . . . . . . . . . . . . . 17
2.2.4 Lower and Upper Bound Theorems . . . . . . . . . . . . . . . 17
2.2.5 Limit Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2.6 Shakedown Analysis . . . . . . . . . . . . . . . . . . . . . . . 23

3 Calculation 26
3.1 2D Simulation for Strip Footing . . . . . . . . . . . . . . . . . . . . . 26
3.1.1 Tresca Material . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.1.2 Coulomb Material . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2 Calculate Loading Condition . . . . . . . . . . . . . . . . . . . . . . . 30

4 Plan and Schedule 35

1
List of Figures

1.1 High Potential Areas for Tidal Resource . . . . . . . . . . . . . . . . . 3

2.1 Different Tidal Power Generating Methods . . . . . . . . . . . . . . . 7

2.2 Primary types of tidal current turbines . . . . . . . . . . . . . . . . . . 9
2.3 Schematic of the sheath system . . . . . . . . . . . . . . . . . . . . . . 11
2.4 Schematic of the anchored system . . . . . . . . . . . . . . . . . . . . 11
2.5 Schematic of the guyed system . . . . . . . . . . . . . . . . . . . . . . 12
2.6 Schematic of the top mounted system . . . . . . . . . . . . . . . . . . 12
2.7 Schematic of the telescopic system . . . . . . . . . . . . . . . . . . . . 13
2.8 The four primary foundations for TCTs . . . . . . . . . . . . . . . . . . 15

3.1 Bearing capacity of strip footing on Tresca material . . . . . . . . . . . 27

3.2 The effects of increasing friction angle on soil bearing capacity . . . . . 28
3.3 Effects of mesh density on FE simulation . . . . . . . . . . . . . . . . 30
3.4 The time varying forces on the turbine . . . . . . . . . . . . . . . . . . 31
3.5 The geometry of a hydrofoil . . . . . . . . . . . . . . . . . . . . . . . 32
3.6 Velocities and forces on one blade element . . . . . . . . . . . . . . . . 33

2
List of Tables

2.1 Properties of traditional methods used for geotechnical stability analysis 19

3.1 Normalized bearing capacity with different friction angles . . . . . . . 28

3.2 The effects of increasing dilation angle on soil bearing capacity . . . . . 29
3.3 design parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.4 Loading condition on the tidal turbine . . . . . . . . . . . . . . . . . . 34

3
Chapter 1

Basis

1.1 Introduction
Tidal current energy, which is caused by the ebb and flood of a tide due to the gravi-
tational pull of the moon,is one of the most attractive and promising renewable energy
sources owing to its advantages relative to other renewables: no land occupation, a
high load factor (water density is 800 times higher than air density), flow predictability
and consequently power production predictability, in-existence of extreme flows (which
might otherwise damage the conversion devices), and etc. are some of the advantages
of tidal stream energy exploitation[2, 3, 6, 9].
A device that uses tidal current by converting kinetic energy into rotational energy to
generate electricity is called tidal current turbine(TCT). The turbine along the rotation
axis direction can be classified into a horizontal axis turbine (HAT) and a vertical axis
turbine (VAT). HAT systems are being studied worldwide[24].
In terms of the installation method, the HAT systems can be divided into two types:
i Floating beneath the surface and anchored to the bottom by means of chains or
cables
ii Rigidly attached to the bottom by means of a structure[21, 26]
The performance of both types HAT systems depends on the distribution of vertical flow
velocity to a great extent[43]. Despite of the hydrodynamic condition, the bottom-fixed
HAT still shows a better performance in coastal engineering, considering the power
transmission and power farm installation. As the bottom-fixed HAT system is well
studied and put into experiment wildly, It’s reasonable to choose it as the research
objective in stability analysis.

1
In geotechnical engineering, stability analysis is used to predict the maximum load that
can be supported by a geostructure without inducing fatigue. This ultimate load, which is
also known as the limit or collapse load, can be used to determine the allowable working
load by dividing it by a predetermined factor of safety[47]. For the purpose of cutting
down the installation cost and extending the stable operating life, to apply the stability
analysis for the TCT foundation in the design procedure is of great importance.

1.2 Motivation
Tidal current energy, as a promising renewable energy, is abundant around the world(see
Figure 1.11). However, due to the relatively recent interest in this type of energy (in
comparison with other renewables such as wind or solar energy), only a few TCTs (Tidal
Current Turbines) have achieved a commercial stage. High installation cost and low
power output efficiency are the two major bottlenecks hindering the commercialization
of TCT technology.The focuses of recent researches on the TCT mainly involve blade
design, TCT farm design and performance of different types of TCT. The methods used
in the stability analysis of the structure are based on the technologies in the design
of offshore wind turbine,whose supporting structures and anchoring systems are quite
similar to those of TCTs. However, due to the high density of seawater, the drag and
lift forces are much larger than in air, causing extensive stress and deflection to the
TCT structure[25]. Meanwhile, neither the huge submarine turbine nor rushing cyclic
tidal current are neglected factors in stability analysis. So, it’s meaningful to apply
shakedown theory,which can deal with the complex multi-loading condition,especially
including cyclic loading, in the stability analysis of TCT foundation.
1www.renewablegreenenergypower.com

2
 Figure 1.1: High Potential Areas for Tidal Resource

1.3 Aims,objectives and outputs

The ultimate goal is to ensure the bearing capacity of Tidal Current Turbine foundation
under the complicated hydrodynamic condition using shakedown theory. This requires
to specify the loading condition on the whole structure caused by the cyclic tidal current
and surface wave which may exerts significant impact on the turbine if the level of wave
movement is high enough. We need to consider the scouring effect around the foundation
as well. Sine the shape of TCT systems various from the working environment, We will
choose several popular types based on the previous researches. The outputs expected is
listed in the following:
Integrated loading condition-It’s a precondition for the stability analysis. Most previ-
ous researchers only consider one or two dominating forces acted on the structure, and
also simplify the periodic loading using constant average value. The integrated loading
condition contains: 1) the forces applied on the turbine by surface wave and current;
2) stress imposed on the supporting structure and foundation caused by cyclical tidal
stream; 3)perturbation due to turbulent flow; 4) asymmetric soil condition around the
foundation created by water scouring; 5) hydrostatic pressure.
Ultimate bearing capacity-To obtain the bearing capacity under common engineering
circumstance is in the great significance rather than get that with extreme loading.
According to shakedown theory, we can observe the response of the TCT foundation
under the integrated loading condition.

3
Safety factors-The calculation of bearing capacity is aim to figure out how the safety
factors contribute to the stability condition. Apart from the environmental factors like
velocity profile of tidal current, we also need to understand the shape factors and their
relations to the ultimate bearing capacity.

1.4 Methodology

1.4.1 Reflection on existing projects

The research is to reflect the current stability analysis projects of offshore foundation with
respect to soil materials, soil-structure interaction, simulating techniques and calculating
algorithms. We can gain great experience from the past projects on stability analysis.
Systematic reflections on these projects can help us sort out all aspects during building
the creep-fatigue model under cyclic lading condition. The aim is to unveil the unsolved
problems, limitations, potential challenges and opportunities of stability analysis projects
of offshore foundation.

1.4.2 Loading calculation

To improve the accuracy of stability analysis, we need essentially obtain the actual loading
condition of the TCT. The structure underwater is sensitive to the environmental effects,
especially with irregular geometry. In general, the computational fluid dynamics(CFD)
program is the first choice in dealing with the fluid-structure interaction problem. How-
ever, a complete simulation of the engineering environment will not only cost a huge
amount of time, and there would be quite difficult to inspect the accuracy of the results,
which is gonna inconvenience the process of stability analysis. In this case, a series of
non-linear external forces caused by water movements with various of frequency are cal-
culated. Alternatively, we will implement the calculation through empirical equations,
which is relatively cost-effective and saving time with sufficient accurate requirement.
In the later of the research, we can also design some simple experiments to amplify the
influence of single environmental variable, to verify the loadings got from the empirical
equations, to observe the response of structure under combining-loading as well.

1.4.3 Finite-element simulation

The finite element method (FEM) is a numerical technique for finding approximate so-
lutions to boundary value problems for partial differential equations. It uses subdivision

4
of a whole problem domain into simpler parts, called finite elements, and variational
methods from the calculus of variations to solve the problem by minimizing an associ-
ated error function. Analogous to the idea that connecting many tiny straight lines can
approximate a larger circle, FEM encompasses methods for connecting many simple
element equations over many small subdomains, named finite elements, to approximate
a more complex equation over a larger domain. The subdivision of a whole domain into
simpler parts has several advantages[40]:
• Accurate representation of complex geometry
• Inclusion of dissimilar material properties
• Easy representation of the total solution
• Capture of local effects
Through the method, we can discretize the integral formations of the geotechnical
stability analysis, which do realize the calculation of large scale and three-dimensional
modal simulation. The computer grogram like ANSYS and ABAQUS developed later
further convenient the modeling and simulating processes in solving geo-problems.

5
Chapter 2

Literature Review

2.1 Tidal Current Turbine

2.1.1 Tidal Energy

Tidal energy is one of the most exciting emerging forms of renewable energy. It is a
form for water power, or hydro-power as it is otherwise known, and when captured it
can produce useful energy such as electricity. Tidal currents, unlike many other forms
of renewable energy, are a consistent source of kinetic energy caused by regular tidal
cycles influenced by the phases of the moon. Intermittency is a problem for wind,
wave and solar power as the sun doesn’t always shine and the wind doesn’t always blow.
These sources of renewable energy often require backup from traditional forms of power
generation. However, the inherent predictability of tidal power is highly attractive for
grid management, removing the need for back-up plants powered by fossil fuels.
Tidal energy is probably the oldest form of Renewable Energy Source which was used
for commercial purposes and for the benefit of mankind. It was used in the watermills
along the shores of England during the Roman age and also used along the shores of
France and Spain. This period was around 900 AD. The water rising from the tides
was impounded in barrages and it was allowed to flow through waterwheels and other
energy capturing devices thus capturing the energy and converting it into useful work
used to grind grains and produce flour. The first tidal mill in the US was built in Salem,
MA in 1635. In recent history it was not until the 1960s that a tidal energy plant was
built in France, on the estuary of La Rance, to capture the tidal energy and produce
electricity. The plant has been operating for the past 50 or so years producing electricity
without missing a tide. The second tidal energy plant was built about 20 years later in

6
1982 in Nova Scotia, Canada.Nowadays,tidal power can be classified into four generating
methods:

(a) Tidal Current Generator (b) Tidal barrage

(c) Dynamic tidal power (d) Tidal lagoon

Figure 2.1: Different Tidal Power Generating Methods

Tidal Current Generator-It makes use of the kinetic energy of moving water to power
turbines, in a similar way to wind turbines that use wind to power turbines. Some tidal
generators can be built into the structures of existing bridges, involving virtually no
aesthetic problems. Land constrictions such as straits or inlets can create high velocities
at specific sites, which can be captured with the use of turbines. These turbines can be
horizontal(Fig 2.1a), vertical, open, or ducted and are typically placed near the bottom
of the water column1.
Tidal Barrage-These barrages utilize the potential energy in the difference in height (or
hydraulic head) between high and low tides. When using tidal barrages to generate power,
the potential energy from a tide is seized through strategic placement of specialized
dams(Fig 2.1b). When the sea level rises and the tide begins to come in, the temporary
increase in tidal power is channelled into a large basin behind the dam, holding a large
1www.tethys.pnnl.gov

7
amount of potential energy. With the receding tide, this energy is then converted into
mechanical energy as the water is released through large turbines that create electrical
power through the use of generators[39]. Barrages are essentially dams across the full
width of a tidal estuary.
Dynamic Tidal Power-It is an untried but promising technology that would exploit an
interaction between potential and kinetic energies in tidal flows. It proposes that very
long dams (for example: 30−50 km length) be built from coasts straight out into the sea
or ocean, without enclosing an area(Fig 2.1c). Tidal phase differences are introduced
across the dam, leading to a significant water-level differential in shallow coastal seas,
featuring strong coast-parallel oscillating tidal currents such as found in the UK, China,
and Korea.
Tidal Lagoon-A newer tidal energy design option is to construct circular retaining
walls embedded with turbines that can capture the potential energy of tides(Fig 2.1d).
The created reservoirs are similar to those of tidal barrages, except that the location
is artificial and does not contain a pre-existing ecosystem. The lagoons can also be in
double (or triple) format without pumping or with pumping that will flatten out the power
output[32]. The pumping power could be provided by excess to grid demand renewable
energy from for example wind turbines or solar photovoltaic arrays. Excess renewable
energy rather than being curtailed could be used and stored for a later period of time.
Geographically dispersed tidal lagoons with a time delay between peak production would
also flatten out peak production providing near base load production though at a higher
cost than some other alternatives such as district heating renewable energy storage. The
proposed Tidal Lagoon Swansea Bay in Wales, United Kingdom would be the first tidal
power station of this type once built2

2.1.2 Offshore Tidal Turbine

Tidal current turbines extract the kinetic energy in moving water to generate electricity.
Tidal current technology is similar to wind energy technology[41]. However there are
several differences in the operating conditions. Under similar conditions water is 832
times more dense than air and the water flow speed generally is much smaller[4]. Since
tidal current turbines operate in water, they experience greater forces and moments than
wind turbines. Tidal current turbines must be able to generate during both flood and ebb
tides and be able to withstand the structural loads when not generating electricity. The
following four methods of tidal current energy extraction are the most common[5]:
2www.renewableenergyfocus.com

8
 Figure 2.2: Primary types of tidal current turbines

• Horizontal axis tidal current turbine- The turbine blades rotate about a hori-
zontal axis which is parallel to the direction of the flow of water, extracting energy
from moving water in much the same way that wind turbines extract energy from
moving air.
• Vertical axis tidal current turbine- The turbine blades rotate about a vertical
axis which is perpendicular to the direction of the flow of water[27].
• Oscillating hydrofoil- It’s also known as stingray Tidal Energy Generator devel-
oped by Engineering Business Ltd. Stingray uses the flow of the tidal stream over
a hydroplane to create an oscillating motion that operates hydraulic cylinders to
drive a motor that, in turn, drives an electrical generator.
• Venturi effect turbine- Different from horizontal axis current turbine, It utilizes
the principle of the Venturi effect as a modification. During the process, the water
is fed through a tapered pipe, which increases the velocity of the water as well as
the energy output.
The technology of axis turbines are more sophisticated then the later ones, which actually

9
are still under small-scale test. Nowadays, the vertical axis tidal current turbine having
its gearbox and generator above sea level does not seem to have any advantage over the
horizontal axis type, since most of which are are designed to be raised out of the water
for repairs, like the SeaGen installed in Strangford Lough in Northern Ireland in 2008.
On the other hand, the gearbox and electric generator portion would be accessible on
the vertical axis tidal current turbine without lifting but a special hoist would have to be
used to raise up the generator to examine the blade section. Normally, it is much safer
to work on ocean current turbines out of the water since the currents are swift and can
be dangerous. Therefore, it’s reasonable for us to choose horizontal axis tidal current
turbine as our research object.

2.1.3 Supporting Structure

As mentioned above, the most obvious additional consideration for TCTs, comparing
with wind turbines, is the marine environment. The additional complexity required
and resistance to corrosion mean that the costs incurred during offshore operations
may prove pivotal in the economic viability of a particular design. The ability of the
structure to facilitate offshore installation, maintenance and decommissioning of the
system is therefore of the utmost importance[37]. A number of supporting structure are
already under development. Different devices are suite to different financial models and
environment conditions. In other words, different supporting structures contribute to the
various loading conditions undertook by the foundations, which do really matters in our
stability analysis. Several concepts of supporting structure are chosen for analysis:
The sheath concept (Fig 2.3) is based on the Marine Current Turbines design[19]. A
tower is installed on the seabed so that it pierces the surface. The rotor and generator
are moved up and down the tower mechanically, which reduces the cost of maintenance
if malfunction occurs, but also gives rise to the effect of corrosion in splash zone. The
electrical conditioning and monitoring equipment are located in the nacelle at the top
of the tower. The installation involves floating the tower out to site, and then locating it
on the seabed. Subsequently the sheath, rotor and nacelle are lifted into position. All
these procedures required accurate positioning of installation vessel. Minor maintenance
operations are undertaken in the nacelle and with the rotor raised. More major operations
involve the removal of the drive train to shore.

10
 Figure 2.3: Schematic of the sheath system

The anchored-buoyant system is based on the proposed SMD Tidal design[28]. The
operating condition is shown on the left of Figure 2.4. The buoyancy of the nacelle gives
tension to the anchor chain and the rotor is held in a downstream position. This simple
mechanism manifests itself a deep water tolerance and self-aligning to flow direction.
However, obviously, the flexible structure makes it vulnerable to fatigue loading. For
maintenance the chain can be released, raising the nacelle to the surface where it is
towed or transported to shore. During commissioning the chains can be installed on the
seabed with a buoy to float one end. The nacelle is then attached to the chain and cable.
The chain is then retracted to submerge the nacelle and probably secured using chain
clamps.

Figure 2.4: Schematic of the anchored system

The guyed tower design (Fig 2.5) uses the buoyancy of the nacelle to tension multiple
chains. Owing to the location of the chains they provide stability to the rotor during
operation, sacrificing the efficiency of capturing the tidal energy. For maintenance the

11
chains can be released, raising the nacelle to the surface where it is towed or transported
to shore. During commissioning the chains can be installed on the seabed with a buoy
to float one end. The nacelle is then lowered using variable buoyancy and attached to
the chains and cable. It is secured in position using chain clamps, probably operated by
divers and finally the positive buoyancy is reinstated to tension the chains.

Figure 2.5: Schematic of the guyed system

The top mounted nacelle concept (Fig 2.6) is similar to the Hammerfest Strom AS
design3. It employs a tower or structure to support the rotor at the required depth. The
tower is installed on the seabed prior to the lowering and connection of the nacelle
and cable. This is achieved using a mechanical locking device and probably divers.
Maintenance is simply the reverse of the latter part of this operation, leaving the tower
in place on the seabed. The nacelle is then towed to shore. This kind of supporting
structure is relatively stiffer, but even minor maintenance requires complete removal of
the nacelle to shore.

Figure 2.6: Schematic of the top mounted system

The telescopic concept (Fig 2.7) is based on a design proposed by Swan-turbines[38]. A

3www.e-tidevannsenergi.com

12
system of telescopic towers is used to maintain the nacelle at the required depth during
operation. In operation the towers are in the contracted position and for maintenance
the sections are extended using buoyancy leaving the nacelle accessible at the surface.
The sections are clamped in the extended position to maintain stability. The installation
procedure involves installing the lowest tower section and subsequently interlocking the
upper sections and the nacelle. Minor maintenance operations are undertaken in the
nacelle and with the rotor raised. More major operations involve the removal of the drive
train to shore. The main disadvantages of telescopic turbine are complexity of telescopic
joint and high requirement of the steel.

Figure 2.7: Schematic of the telescopic system

Besides, the turbine radius and the cross section of the tower always play an important
role in the design of TCTs[35].

2.1.4 Foundations and Anchoring System

The tidal current energy system developers share a common challenge. Each device
needs to be held on station in a tidal flow in a practical and cost-effective way. The
foundations or mooring systems required can form a significant part of the life-cycle cost
of the system since they contribute to the capital cost and the ease of installation and
maintenance of the device. There are many methods of solving this problem ranging
from well-established solutions to new and innovative solutions.
Offshore structures are subjected to severe environmental loads from waves, wind,
current and possibly ice and earthquakes. Strict requirements are set for the optimum
performance of their foundation systems and the foundation design is a key activity in
the overall engineering process of these structures. The foundations must be designed
for the expected cyclic load history. The capacity under cyclic loading may be higher or
lower than the monotonic capacity, depending on the cyclic degradation[1].

13
Various types of structures and foundation systems are used such as piled jackets, jackets
with caissons (bucket foundations), jack-up rigs on spud-cans or mats, floating structures
moored to different types of anchors and steel or concrete gravity base structures with
or without skirts. In this section, we will introduce to you several primary foundation
types , comparing them in the aspects of bearing capacity, installation, stiffness and
soil-structure interaction.
Monopile foundation-The monopile, which consists of a foundation pile and a transition
piece, is normally a highly efficient type of foundation and is therefore at present the
most popular option at water depths not exceeding 20−25m (See in Fig 2.8 a). With
high strength steel, it can also applied in the depth of 35m[44]. One monopile can have
a weight of up to 700 tonnes of steel. The standard method of installation of monopile
is to lift or float the structure into position and then drive the piles into the seabed using
either steam or hydraulic powered hammers. With high and varying steel prices the
material costs are an important part of the total structure costs. The total construction
costs are around 1500 per tonne of monopile foundation[29].
Gravity base foundation-This type of foundation is normally made of concrete, which
can be constructed with or without small steel or concrete skirts. The ballast required to
anchor the foundation consists of sand, iron ore or rock filled into the base of the founda-
tion with adjustments in the designed base width to suit the soil conditions. The design
will include a central steel or concrete shaft for transition to the supporting structure.
For most locations the foundations will require some form for scoring protection (See in
Fig 2.8 b). Due to the low material costs the construction of the concrete GBF is around
200 per tonne which is cheaper when compared to a steel monopile[29].
Suction bucket foundation The Bucket foundation combines the benefits and main
proven aspects of a gravity base foundation, a monopile and a suction bucket into one
product. It includes a patented cost efficient installation system which controls the
vertical alignment of the total foundation as it sucks itself into the seabed reducing total
installation time significantly. Being lowered upside-down into the seabed, a quicksand
is created around the rim of the bucket with a pipe that runs up through the stem above
sucking water out of the bucket. When it firmly sinks into the seabed, the pump is turned
off. Suction buckets have been extensively used as anchors, principally in clay, for a
range of water depth form 10m to 20m typically[23].
Tripod foundation-The tripod foundation is considered to be a relatively lightweight
three-legged steel jacket compared to a standard lattice structure. Under the steel central
column, which is below the turbine, there is a steel frame which transfers the forces
from the tower into the three steel piles. Piles are installed at each leg position to anchor
the tripod to the seabed. The three piles are driven 10−20 m into the seabed. The
tripod can also be installed using suction buckets[45]. The advantages of the tripod are

14
the suitability for greater water depths, and a minimum of preparations required at the
site before installation. Erosion is not normally a problem associated with this type of
foundation. They go on to cite the foundations potential to save costs compared to a
more complex jacket design. The effect of scour can be less significant when compared
to the monopile foundation(See in Fig 2.8 c,d).

(a) Monopile foundation (b) Gravity base foundation

(c) Suction bucket founda- (d) Tripod foundation

tion

Figure 2.8: The four primary foundations for TCTs

Optimum foundation selection is a function of the variables of water depth, turbine size,
anchoring system and location conditions. These facts, especially the soil condition, have
great influence on the stability of the foundation. The monopile and gravity foundation
will be investigated first because of their simple structures and anchoring system. The
stability of suction bucket foundations is also worthy studying since they are popular in
the offshore engineering.

15
2.2 Stability analysis
The problem of soil mechanics are often divided into two distinct groups: the stability
problems and the elastic problems. They are treated in two separate and different ways.
The stability problems deal with the conditions of ultimate failure of a mass of soil.
The elastic problems on the other hand dear with stress or deformation of the soil when
no failure of the soil is involved[11]. Broadly speaking, there are four main methods
for performing geotechnical stability analysis: limit equilibrium, limit analysis, slip-
line methods, and the displacement finite-element method. In this research, focus is
concentrated on the shakedown analysis, while limit analysis, which is more traditional
and sophisticated, is used for comparison.

2.2.1 Slip-line method

The slip-line equations were first to be derived by Kotter for the case of plane deforma-
tions. Prandtl was the first to obtain an analytical closed form solution to these equations
for a footing on a weightless soil. Although the method has been developed for some
simple two dimensional cases, the neglect of the stress-train relationship of the soil,
which must be satisfied for a valid solution, hinders the wild application of the slip-line
method. Meanwhile, it is not well suited to the development of general -purpose software
which can deal with a wide variety of practical problems.

2.2.2 Limit equilibrium

Limit equilibrium is the oldest method for performing stability analysis, and was first
applied in a geotechnical setting by Coulomb[13]. In its most basic form, this approach
presupposes a failure mechanism, and implicitly assumes that the stresses on the failure
planes are limited by the traditional strength parameters c and ϕ, which denotes the
cohesion yield stress and friction angle respectively. The chief advantages of the limit-
equilibrium method are its simplicity and its long history of use, which have resulted in
widely available software and extensive collective experience concerning its reliability.
Its main disadvantage, on the other hand, is the need to guess the general form of the
failure surface in advance, with poor choices giving poor estimates of the failure load[47].
Despite of shortcomings, the limit equilibrium method has still been implemented in
geotechnical problems, particularly for slop stability analysis[18, 10, 12].

16
2.2.3 Displacement finite-element analysis

As a result of the rapid evolution of powerful user-friendly software, displacement finite

element analysis is now widely used in geotechnical practice, not only for the prediction
of deformations, but also for the prediction of stability. This method is very general, and
can accommodate advanced constitutive models that incorporate non-associated flow,
heterogeneity, anisotropy, and work/strain-hardening and softening. In addition, robust
procedures are available for modelling interface behaviour, soil-structure interaction and
large deformations, as well as fully coupled consolidation and dynamics. When it is
used to predict stability under static loading, displacement finite-element analysis can
be used in two different modes[47].
(a) The loads are applied in increments until the deformation response indicates that
a state of collapse has been reached. The approximate ultimate load so obtained
furnishes a safety factor in terms of force, not strength, and requires a complete
simulation of the load-deformation response[48, 16].
(b) Successive analyses with reduced strengths are conducted until equilibrium can no
longer be maintained[20, 14]. This approach, known as strength reduction analysis,
involves monitoring deformations at specified control points in the soil, and gives
the safety factor in terms of strength, much like the safety factor, that is computed
in traditional slope stability calculations using limit equilibrium.
Unlike the methods discussed previously, stability analysis with the displacement finite-
element method requires not only the conventional strength parameters, but also the
deformation parameters, like Poisson’s ratio and shear modulus. Since a displacement
finite-element solution satisfies equilibrium and the flow rule only in a ’weak’ sense
over the domain, the quality of the resulting collapse load prediction is often critically
dependent on the mesh adopted. In addition, the accuracy of the limit load can be affected
by the number of load steps used in the analysis[46]. In this research, we are goona to
reproduce the effects on the simulation results of the mesh density, load increments, and
the locking problem might occur.

2.2.4 Lower and Upper Bound Theorems

To obtain a exact value of fatigue loading is impossible technically, usually we set

boundaries around the true value via some numerical methods. Before we go to the next
stability analysis method, it is necessary to introduce two basic theorems used in the
limit analysis as well as shakedown analysis later.

17
Lower Bound Theorem

For the lower bound theorem, a statically admissible stress field is defined as one which
satisfies the equilibriums and the boundary conditions, with nowhere violating the yield
criterion. If all changes in geometry occurring in collapse are neglected, a statically
admissible collapse load factor is always less than or equal to the exact collapse load
factor (mcs ≤ mc ). The equality sign is valid only when the statically admissible stress
field is the true stress field. In other words, the load factor derived from then statically
admissible stress field is a lower bound on the true collapse load factor. This theorem
can be proven by considering the expression below:
∭
(σi j − σisj )dεi j dV
V

Where the true stress and strain fields at plastic collapse are denoted by (σi j , εi j ) and
the statically admissible stress filed is denoted by σisj which is in equilibrium with the
load factor mcs . Since the actual work done in a given plastic train rate is greater than
fictitious work done by an arbitrary state of stress not exceeding the limit yield, the
above expression is consequently non-negative. Then, by the principle of virtual work,
we have:
∭ ∬
(σi j − σisj )dεi j dV = (mc − mcs ) T0i dui dS ≥ 0 (2.1)
V ST
Where the surface traction is denoted by mcs T0i , which is in equilibrium with the statically
admissible stress field. With calculation, the equation (2.1) can lead to the lower bound
theorem: mcs ≤mc .

Upper Bound Theorem

Different from the lower bound theorem, the upper bound theorem assumes a deforma-
tion mode (or velocity field) that satisfies: (a)velocity boundary conditions; (b)strain
and velocity compatibility conditions. This velocity field is termed a kinematically
admissible velocity field. The loads, determined by equating the external rate of work
to the internal rate of dissipation in the assumed velocity field, are not less than the
actual collapse load. The method to prove the upper bound theorem is quite similar with
that in former section. We denote the true stress field at plastic collapse and the the
kinematically admissible strain rate field by σi j and dεikj respectively. The stress field
σikj is associated with the kinematically admissible strain rate field. Then, combining
the principle of virtual work and maximum plastic work, we can have:

18
 ∭ ∬
(σikj − σi j )dεikj dV = (mck − mc ) T0i dui dS ≥ 0 (2.2)
V ST

After simplification, the upper bound relationship (mck ≥ mc ) is obtained.

2.2.5 Limit Analysis

Limit analysis is based on the plastic bounding theorems developed by Drucker et

al.[17], and assumes small deformations, a perfectly plastic material, and an associated
flow rule. The last assumption, which is often termed the normality rule, implies that
the plastic strain rates are normal to the yield surface. For this type of plasticity model
it is necessary to work with velocities and strain rates, rather than displacements and
strains, as the latter become undefined at collapse. Conventional limit analysis has the
intrinsic advantage of providing solutions that bound the collapse load from above and
below, but it is restricted to the use of simple soil models and is often difficult to apply
in practice. Notwithstanding the limitations that stem from the assumption of a simple
perfectly plastic material model, the ability of the limit theorems to provide rigorous
bounds on the collapse load is one of their great attractions. The table 2.1 summarises the
key features of the upper-bound(UB) limit analysis, lower-bound(LB) limit analysis and
displacement finite-element(FE) approaches for assessing geotechnical stability.

Property UB limit LB limit Displacement

analysis analysis FE analysis
Assumed failure mechanism? Yes - No
Equilibrium satisfied everywhere? - Yes nodes only
Flow rule satisfied everywhere? Yes - integration
points only
Complex loading possible? Yes Yes Yes
Complex boundary conditions? Yes Yes Yes
Complex soil models possible? No No Yes
Coupled analysis possible? No No Yes
Table 2.1: Properties of traditional methods used for geotechnical stability analysis

Although the limit theorems can be applied in an analytical setting to give useful bounds
for simple problems, discrete numerical formulations provide a more general means of
harnessing their power. In particular, finite-element limit analysis formulations have
evolved rapidly in recent years, and are now sufficiently developed for large-scale prac-
tical applications in geotechnical engineering. Finite-element formulations of the limit
theorems inherit all the advantages of the finite -element method, and can model complex

19
geometries, layered soils, anisotropy, soil-structure interaction, interface effects, discon-
tinuities, complicated loadings, and a wide variety of boundary conditions. The success
of this approach, however, hinges on the development of formulations and solution
algorithms that are robust, efficient and extendible to three dimensions.

Finite-element lower-bound analysis

Lysmer was an early pioneer in applying finite elements and optimisation theory to com-
pute rigorous lower bounds for plane-strain geotechnical problems, in which cases the
Mohr-Coulomb yield criterion was chosen as the failure criterion. At that time, to utilize
the Mohr-Coulomb yield function in its native form is of great inefficiency, sine this
progress produces a huge number of high order equations that will take lots of time for
calculation. To avoid this type of constraint, Lysmer linearised the yield surface using
an internal polyhedral approximation that replaced each non-linear yield inequality con-
straint by a series of linear inequalities[31]. The accuracy of the resulting linearisation
can be controlled by varying the number of sides in the polyhedral approximation, with
the highest accuracy being obtained at the cost of additional constraints and increased
solution times. The improvements have been made by lots of scientists in the different
aspects of the lower-bound method, but the essential procedures are the same:
• First we need to choose the element type for the case and define the shape function,
whose applicability will directly decide the degree of approximation to the reality
and the complexity of calculation. In simple geotechnical problems, the three-
node triangular and the linear shape function are common choices, which are
gonna to be sued in the following procedures.
• To ensure the stresses satisfying the equilibrium conditions at every point in an
element, we need impose the constraints in equation (for a two-dimensional case)

∂σx ∂τxy ∂σy ∂τx y

+ + h x + g x = 0; + + h y + gy = 0 (2.3)
∂x ∂y ∂y ∂x
Where the h x , h y and g x , gy denote the unknown body fore and fixed body fore
respectively.
• In order to permit statically admissible discontinuities at the edges of adjacent
triangles, it is necessary to enforce additional constraints on the nodal stress. The
basic form of discontinuity equilibrium can be illustrated through the case with
two triangles, a and b, which share a side d defined by the nodal pairs (1,2) and
(3,4):
σn1 σ σn3 σ
 a  a  a  a
= n2 = n4 (2.4)
τ1
a τ2
a τ3a τ4a

20
 • Combine the stress boundary condition. To satisfy boundary equilibrium, the
stresses for any boundary node must match the prescribed surface tractions, which
are commonly defined in terms of normal and tangential components along a
boundary edge.
• The last constraint is the yield function. Not only the type of yield criterion,
the form of the yield function does matters in the whole process as well. For a
linear programming problem, it is necessary to linearise the yield criterion to a
linear function with the unknown stresses. If we choose a non-linear optimisation
algorithm, like the two-stage quasi-Newton slover[30], the yield function can be
utilized in its native form.
• At the end, a function need to be specified to connect the stresses to the bearing
capacity. In many applications, the objective function to be maximised corre-
sponding to a force acting along the boundary of the domain (the collapse load).
Then the problem of finding a statically admissible stress field which maximize
the collapse load for a two dimensional case may be state:

Minimize cT σ
Subject to [A1 ]σ ≤ b1 (2.5)
[A2 ]σ ≤ b2

Where [A1 ] is the matrix of yield constraints and [A2 ] is a matrix of all the other
equality constraints. The cT , b1 and b2 are vectors of constant. Solving this
optimization problem, we will get the bearing capacity.

Finite-element upper-bound analysis

The upper bound theorem of classical plasticity theory, which assumes a perfectly plastic
soil model with an associated flow rule, is a useful tool for predicting the stability of
problems in soil mechanics. It states that the power dissipated by any kinematically
admissible velocity distribution ui that can be equated to the power dissipated by the
external loads, and so enables a strict upper bound on the true limit load to be deduced.
Mathematically, this can be shown by equating function (2.6) and (2.7).
∫ ∫
WÛ = Pint − T
t udA − gT udV (2.6)
At V
∫ ∫
Pext = q udA +
T
hT udV (2.7)
Aq V

21
Where q, t are surface stresses and g, h are body forces. The Pint is the plastic dissipation,
defined by ∫
Pint = σT εÛ p dV (2.8)
V

A kinematically admissible velocity field is one which satisfies compatibility, the flow
rule and the velocity boundary conditions. To handle the velocity discontinuities, an
efficient formulation of the finite-element upper-bound method has been introduced by
Lyamin and Sloan[30]. By following it, in the case of plain strain problem in cohesive
soil, the upper bound limit load is found by using equations (2.6) and (2.7), noting that
t=g=h=0 for this case, to minimise Pint . For the sake of highlighting the differences to
the lower bound limit analysis, the three-node triangle is chosen as the base element.
Each node i is associated with a vector of two unknown velocities, and each element e is
associated with a vector of three unknown stresses and an unknown non-negative plastic
multiplier rate λ.
Û The main steps is followed:

• In the finite-element upper-bound formulation, the objective function corresponds

to the internal rate of energy dissipated by plastic shearing less the energy expended
by the fixed external forces. The key point here is to discretize the the integral
formation of the internal power dissipation. Noting that the stresses and plastic
strain rates are constant over each element, and summing over all the elements,
the internal power dissipation may be written as
∫ Õ
Pint = σT εÛ p dV = (σT εÛ pV)e (2.9)
V e

This This quantity can be evaluated conveniently by observing that the plastic
strain rates are related to the nodal velocities by the strain–displacement relations

εÛ p = B e ue (2.10)

where [B] is the same constraint matrix derived from the linear shape function
and compatibility equations with lower-bound limit analysis, sine we choose the
same element type here.
• For a associated flow rule, a kinematically admissible velocity field should satisfy
the following condition

εÛ p = λ∇
Û f (σ e ), λÛ ≥ 0, λÛ f (σ e ) = 0 (2.11)

• Across the discontinuity, velocity jumps can occur in the normal and tangential
directions, so that the velocities can potentially differ at nodes that share the same

22
 coordinates. This is a quite different condition from the lower-bound limit analysis.
The flow rule in discontinuity can be defined as
  λÛ ∂ f 
∆un ∂σn
= , λÛ ≥ 0, λÛ f (σn, τ) = 0 (2.12)
∆us λÛ ∂∂τf

Where the ∆un and ∆us are the normal and tangential velocity jumps at a nodal pair
(i,j). For convenience, They can be expressed in terms of the Cartesian velocity
jumps through the equation 2.13.
ij ij
cos β sin β
  
∆un ∆u x
= (2.13)
ij
∆us − sin β cos β ij
∆u y

• Combining the velocity boundary conditions and yield condition, the upper-bound
non-linear optimisation problem can be expressed as

Minimize σT Bu − cT u
e
Subject to B ue = α∇ Û f (σ e )
αÛ ≥ 0
(2.14)
αÛ f (σ e ) = 0
[A]u = b
f (σ e ) ≤ 0
e Í e
where B is the element compatibility matrix, B = B is a global compatibility
matrix, σT Bu is the power dissipated by plastic shearing in the continuum and
discontinuities, cT u is the rate of work done by fixed tractions and body forces, αÛ
is the plastic multiplier times the area for element e, f (σ e ) is the yield function
for element e, [A] is a matrix of equality constraint coefficients, and b is a known
vector of coefficients.

2.2.6 Shakedown Analysis

When an elastic-plastic structure is subjected to a variable loading programme between

extreme values, its behaviour may fall into one of the following types[51].
(1) If the structure is loaded cyclically in tension and compression and also the loading in
both directions produces plastic deformation, then it is likely that fracture will occur
as the result of plastic deformations alternating in sign. This is called alternating
plasticity or low-cycled fatigue.

23
(2) If the structure is loaded cyclically between two extreme values and also the extreme
values are large enough, then it is likely that each loading cycle will produce plastic
deformation. This progressive plastic flow is likely to lead to unacceptably large
deformation after a large number of cycles, which is also known as incremental
fatigue.
(3) If the structure is loaded between two extreme values and also the extreme values
are sufficiently small, then it is likely that at early stage of the loading each cycle will
produce plastic deformation, and therefore continuing loading produces no further
plastic flow. If this happens, then the structure is said to have reached a state of
shakedown, at which the structure will behave in a purely elastic manner.
To overcome the potential difficulties of applying the shakedown concept to engineering
design, we can make use of two fundamental shakedown theorems (seminar to the lower
and upper bound theorems of plastic collapse) developed by Melan and Koiter.

The lower bound shakedown theorem

The lower bound shakedown theorem states that if any self-equilibrated residual stress
field %i j can be found, which, when combined with the elastic stress field produced by
the applied loads, dose not violate the yield condition anywhere, then shakedown will
occur. If the applied load is denoted by λ p0 (where p0 may be conveniently set as the
unit pressure in the actual calculation), and λ is a dimensionless scale parameter, then
all the induced elastic stress components are also proportional to λ. Melan’s shakedown
theorem hence demands

f (λσiej + %i j ) ≤ 0 (2.15)
Where λσiej is the elastic stress field due to the applied pressure p0 , and f (σi j ) = 0 is
the yield condition for the material. The largest value of λ obtained by searching all
possible self-equilibrated residual stress field, %i j , will give the actual shakedown limit
ps = λs p0 .

The upper bound shakedown theorem

The upper bound theorem can be stated as follows: an elastic-perfectly plastic structure
will not shakedown for given extreme load values if any kinematically admissible plastic
strain rate cycle εÛikj and any external loads T0i within the prescribed limits can be found
so that ∫ t ∬ ∫ t ∭
dt T0i uÛi dS >
k
dt σikj εÛikj dV (2.16)
0 ST 0 V

24
Which provides an upper bound for the shakedown limit.

25
Chapter 3

Calculation

3.1 2D Simulation for Strip Footing

Bearing capacity of the footing is often calculated by the superposition method suggested
by Terzaghi[49], in which the contributions of individual bearing capacity factors are
summed. Bearing capacity of a strip footing (a plane strain situation) may be expressed
as

quilt = cNc + qNq + γaNγ (3.1)

in which quilt is the ultimate bearing pressure; c is the cohesion of the soil; γ is the
unit weight of soil; q is the surface surcharge; a is the half width of the footing; Nc ,
Nq , and Nγ are bearing capacity factors representing the effect due to c, q and γ,
respectively, and are the functions of friction angle (φ) of the soil. In this research,
FEM is the key point for the simulation. So it’s reasonable to practise the classical
problems using FE program ABAQUS, which is gonna to be used to simulate limit and
shakedown analysis. Incremental-iterative type FEM has been used by several authors in
the computation of bearing capacity of surface footings. Footings on a purely cohesive
soil[50, 34], cohesive frictional soil[15] and cohesive frictional ponderable soil[8] have
been reported. As following, we will derive the bearing capacity of surface strip footing
in Tresca and Coulomb material using ABAQUS.

26
3.1.1 Tresca Material

In the modeling, soil is idealized as an elastic-perfectly plastic material satisfying the

Mohr-Coulomb criterion. Typical elastic constants are assumed: Young’s modulus of
soil E = 2 × 107 k N/m2 ; Poisson’s ratio ν = 0.3, while cohesion c = 1 × 104 k N/m2 ;
unit weight of soil γ = 20k N/m3 ; The Young’s modulus of the surface footing E =
3 × 1010 k N/m2 , which is close to the property of reinforced concrete. The half width
of the strip footing b=1.5m, while the scale of the soil in two dimensional is 10 × 20m2 .
For the Tresca material, the frictional angle equates to zero. Using the limit equilibrium
method, an exact solution has been found by Prandtl for the case of a strip footing on the
frictionless soil surface, showing below

quilt = (2 + π)c ≈ 5.14c (3.2)

By calculation, the bearing capacity in our model should approximate to 5.14×104 k N/m2 .
Using the three-node triangle element in the modeling, we obtained a modified curve
showing the relationship between the stress and strain of the soil under the footing.

Figure 3.1: Bearing capacity of strip footing on Tresca material

Where the safety factor is a normalized bearing stress divided by the cohesive stress.
The result of the simulation is approximate to the value 5.19, with a 0.8% error to the
exact value 5.14, which is negligible.

3.1.2 Coulomb Material

In this section, the friction angle φ and dilation angle ψ are being studied. Angle of
internal friction for a given soil is the angle on the graph (Mohr’s Circle) of the shear

27
stress and normal effective stresses at which shear failure occurs. It’s a parameter to some
extent represent the bearing capacity of the soil. Commonly the value of the friction
angle would in the range of (15,45)[7, 36]. The angle of friction depends on the angle
of dilation, which in turn depends on density and pressure[22]. The angle of dilation
controls an amount of plastic volumetric strain developed during plastic shearing and is
assumed constant during plastic yielding. The value of ψ = 0 corresponds to the volume
preserving deformation while in shear. Clays (regardless of over-consolidated layers)
are characterized by a very low amount of dilation (ψ ≈ 0). As for sands, the angle of
dilation depends on the angle of internal friction. For non-cohesive soils with the angle
of internal friction φ > 30◦ the value of dilation angle can be estimated as ψ = φ − 30◦ .
A negative value of dilation angle is acceptable only for rather loose sands. In most
cases, however, the assumption of ψ = 0 can be adopted. Keeping the dilation angle
equates to zero, we first observe the effects of increasing friction angle on soil bearing
capacity. For comparison, we obtained several ultimate bearing capacity with different
friction angles using Prandtl solution. The results are showing in the table 3.1.

Table 3.1: Normalized bearing capacity with different friction angles

Friction angle (◦ ) 0 5 10 15 20 25 30
Normalized qult 5.14 6.49 8.35 10.98 14.83 20.72 30.14

In FE modeling, we use the modified quadrilateral element instead of the triangle one,
which turns out to be more accurate and easy to converge. The reduced integration
algorithm is applied to decrease to the number of integration point in a quadrilateral
element from 4 to 1, which makes the calculation efficient. From the results shown in
the Fig 3.2, all the curves represent the characteristic of complete convergency.

Figure 3.2: The effects of increasing friction angle on soil bearing capacity

28
Compare the simulation results with those from Table 3.1. By the increase of the friction
angle, the ultimate bearing capacity of the soil grows significantly. And specially, the
increment of the vale of bearing capacity also rises during each growth of the angle
of internal friction. Meanwhile, we need to notice that the error occurred between the
results when friction angle reaching 25◦ is quiet large than the formers. This is because
a non-associated flow rule with ψ = 0 gives a value lower than its limit solution[33].
Due to the unspecified mechanism of dilation angle in the soil, it is commonly chosen
by references to some experiments. In order to figure out the relationship among the
friction angle, dilation angle and bearing capacity, we set a series of models by FEM.
The models are divided into three groups, the friction angle is fixed at 10◦ , 20◦ , 30◦ for
each group respectively. In each group, the dilation angle changes from 0 by φ/4 every
increment. Keeping the other parameters same with the former cases, the results are
showing in the Table 3.2.

Table 3.2: The effects of increasing dilation angle on soil bearing capacity
Ultimate bearing capacity with different dilation angles (k N/m2 )
Friction angle ψ = 0 ψ = φ/4 ψ = φ/2 ψ = 3φ/4 ψ=φ
10◦ 83822 84229 84506 84458 84137
20◦ 143740 149959 151885 150359 149175
30◦ 297039 304863 307220 306520 304329

Observing the datum in the table, It’s clear that the larger the friction angle is, the bigger
the increment of the ultimate bearing capacity will be when ψ ≤ 3φ/4. If the friction
angle is fixed, the value of bearing capacity will increase by the growth of the angle
of dilation when ψ ≤ φ/2. After reaching the peak, the curve of the bearing capacity
would vibrating slightly around the peak value. In finite element analysis, mesh density
is a critical issue which closely relates to the accuracy of the finite element models
while directly determines their complexity level. Theoretically, densifying the meshing
condition will bring out much more accurate result. Nevertheless, this process will
take more time for calculation, which might be unpractical for large scale model. The
appropriate mesh density therefore is important in building the model. The adaptive
mesh refinement methods will be discussed later, we will first show how the mesh density
will effect the simulation procedure and the calculation result for former cases.

29
 Figure 3.3: Effects of mesh density on FE simulation

3.2 Calculate Loading Condition

As introduced in chapter 1, the loading condition on the tidal turbine is a result of several
environment factors. The semi-empirical formula is not feasible for structure stability
analysis where high accuracy is required. Nowadays, the dynamic loads on the tidal
turbine are commonly calculated using computational fluid dynamics method by FE-
software[25], or making use of the rainflow counting algorithm by Matlab codes[42], etc.
Since the experiment is high cost and limited to the specific engineering environment,
this method is not the first choice but usually carried out as a reference to compare with
the numerical results.
To facilitate the progress of the research, we decide to start with the results from the
scale-experiments tested at Marintek, Norwegian Marine Technology Research Institute,
in Trondheim[42], in which the dimensioning loads for NTNU’s 1 MW reference tidal
turbine are calculated due to the vertical no-slip current velocity profile, wave-current
interaction, the horizontal and vertical velocity component due to the waves, turbulence
and the tidal wave. In terms of the turbine, the time varying forces calculated (shown
in Fig 3.4) are the thrust force T, the main shaft torque Mx , and the thrust induced pitch
moment My , and yaw moment Mz .

30
 Figure 3.4: The time varying forces on the turbine

Where the supporting structure is lattice, which is difficult to handle when we consider
the loading condition on the foundation. This problem can be solved by substituting it
with the monopile supporting structure and the values of the forces on the turbine won’t
be affected. Except the forces on the turbine, we still need to calculate the loadings on
the supporting structure, monopile, for this case. The details of the design parameter for
calculation is listed in the Table 3.3.

Table 3.3: design parameters

Number of turbine blades Nblade 2 -
Turbine Diameter Dt 21 m
Depth of turbine blade tip in the upper pos W Dup 8 m
Distance from blade tip in lower pos to seabed Ni 7 m
Density of water ρ 1028 kg/m3
Turbine speed ωt 1.67 rad/s
Tidal wave period Ttidal 12.42 h
Current velocity at design point Vdesign 2.5 m/s
Diameter of the monopile Dm 1.5 m

In general, the loading condition of the structure in the time-varying velocity field can
be obtained by FE-simulation using FLUENT. However, to match up the progress of
stability analysis, we can simply use the Equation 3.3 to calculate the loading on the
monopile by considering a constant current velocity profile.

1
FD = CD AD ρU 2 (3.3)
2

31
Where CD is the coefficient of flow resistance, 0.7 for this case; AD denotes the projected
area perpendicular to the direction of the current velocity which is 26m2 . The density
of the flow is given in the Table 3.3. It is reasonable to choose the current velocity at
design point as the constant value of the current. The calculated force on the monopile
is 58kN and the induced moment on the foundation is 497k N · m.
Besides, the loadings on the turbine are far more complicated. First, we need to observe
the loaded objective. The cross-section of a turbine blade is shaped like a hydrofoil. The
flow meets the leading edge of the hydrofoil. The angle of attack, α, is the angle between
the relative flow, W, and the chord line. The cord line is the straight line connecting the
leading and the trailing edge of the airfoil and has a length, Lc . The mean camber line
is located halfway between the upper and lower surfaces of the airfoil. A hydrofoil with
its geometry is shown in Fig 3.5.

Figure 3.5: The geometry of a hydrofoil

The undisturbed flow, V, encountered with the hydrofoil can be divided into two com-
ponents: U is the rotational speed of the turbine blade, denoted U = ωt r, where ωt is
the angular turbine velocity and r is the position along the blade; W is the relative flow.
This is a mathematical treatment and shown in Fig 3.7. In reality, the flow is split into
two separate streams over the hydrofoil. The flow velocity increases over the convex
side of the airfoil resulting in a lower pressure compared to the concave side. The upper
side of the hydrofoil is the suction side and the lower side is the pressure side. The
pressure difference between suction and pressure side of the airfoil generates a lift force,
FL , perpendicular to the relative flow. Friction between the flow and the airfoil surface
is the main reason for the drag force, FD , a force which is parallel to the flow. The size of
the lift and drag forces varies with the angle of attack. The torque of the blade element,
M’, will act along the cord line at a distance of Lc /4 from the leading edge, parallel to
the rotational speed of the blade. The thrust force, T’, is perpendicular to the rotational
speed.

32
 Figure 3.6: Velocities and forces on one blade element

Then we can calculate the forces by the Blade element theory. The theory is described
for an infinitely long wing with blade elements of a length dr. The blades are divided
into N sections and the velocities on one blade element expressed with respect to a, a’
and V1 , where the a is the axial induction factor; a’ is the rotational induction factor and
V1 is the far upstream velocity of the turbine. The thrust and torque on one blade element
for a thickness dr with Z numbers of blades can be expressed as follows:

1
dT = Ca ρW 2 Lc dr Z (3.4)
2
1
dM = Cr ρW 2 Lc dr Z (3.5)
2
where the are Ca and Cr the axial and rotational force coefficient:

Ca = CL cos ϕ + CD sin ϕ (3.6)

Cr = CL sin ϕ − CD cos ϕ (3.7)

By expressing the relative velocity, W, with respect to a, a’ and V1 inserting this in
equation 3.8 and 3.9 the new expressions for dT and dM are:

1 V12 (1 − a)2
dT = Ca ρ Lc dr Z (3.8)
2 sin ϕ2
1 V1 (1 − a) rω(1 + a0)
dM = Cr ρ Lc dr Z (3.9)
2 sin ϕ cos ϕ

33
Where CL and CD are variable from the a, which can be found in the references. By
summarizing over the whole blade we can obtain the total thrust force and shaft moment
on the turbine. The yaw moment and pitch moment induced by thrust forces on blades
can be calculated via multiplying dT by the distances to y and z axis respectively. V1 is
varying with the water depth and time, which is effect by the wave, tidal and turbulence.
The tidal current is inducing a vertical no-slip velocity profile. Because of seabed
friction and water viscosity the profile will vary with the water depth. The vertical
velocity profile from the current, not affected by the waves, can be be found with a good
approximation using the power law:

WD + z β
Vz = V0 ( ) (3.10)
WD
Where WD is the total water depth in meters, z is the depth from the mean free sea water
surface in meters, negative downwards, and β is equal to 1/7. If the wave is considered,
the sea surface will oscillate above a mean level. The vertical velocity profile will be
stretched or compressed, so that the surface velocity component is held constant. This
phenomenon is called the Wheeler method. A wave crest will stretch the vertical profile
and the trough will compress it. The depth, z, becomes zs , which is the stretched vertical
coordinate. zs is defined:

η
zs = z(1 + )+η (3.11)
WD
Where η is the sea surface elevation. -WD < z < 0 and -WD < zs < η, zs > z for η > 0
and zs < z for η < 0. Considering the time-vary velocity field, the forces on the turbine
are listed below:

Table 3.4: Loading condition on the tidal turbine

Loading Maximum (kN or k N · m) Minimum (kN or k N · m) Cyclic period
Thrust 940 -20 6.21h
Shaft torque 1041 -40 6.21h
Yaw moment 594 -597 3.77s
Pitch moment 1040 -1041 3.77s

It’s necessary to point out that the yaw moment and pitch moment not only oscillate
during the turbine period, given in the Table 3.4, the magnitude of both moments will
vibrate in every tide wave movement. It means the profiles of the yaw moment and
pitch moment look like two superimposed sin waves with a phrase difference of T/2.
The curves of the shaft moment and thrust can be approximate to the simple sinusoidal
waves.

34
Chapter 4

Plan and Schedule

In the next academic year, we will focus on the research about how to apply the shakedown
theory on the stability analysis of the tidal turbine foundation under cyclic loading. Again,
like what we have done before, we can start with two-dimensional cases and then expend
the method into three-dimensional ones. The brief objectives are listed below.
• Look through the researches of the shakedown theory so that we can make a clear
view on the development of the theory.
• Compare the exiting methods for applying shakedown theory and find a feasible
method for the case of Tidal current turbine foundation.
• Specify the parameters of the soil and the connection condition of between the
soil and foundation.
• Program using finite-element method corresponds to the shakedown theory and
implement it in the existed model.
• Simulate and analysis the results, highlight the effects on ultimate bearing capacity
of different cyclic loadings as well.
• Improve the accuracy by reforming the optimization algorithm.
• Obtain the lower and upper bounds by using limit analysis so that we can compare
with the results from the shakedown analysis.

35
Bibliography

[1] L Andresen, HP Jostad, KH Andersen, and K Skau. Finite element analyses in

offshore foundation design. In The 12th International Conference of International
Association for Computer Methods and Advances in Geomechanics. Goa, pages
3247–3262, 2008.
[2] AbuBakr S Bahaj. Generating electricity from the oceans. Renewable and Sus-
tainable Energy Reviews, 15(7):3399–3416, 2011.
[3] LS Blunden and AS Bahaj. Initial evaluation of tidal stream energy resources at
portland bill, uk. Renewable Energy, 31(2):121–132, 2006.
[4] IG Bryden, T Grinsted, and GT Melville. Assessing the potential of a simple tidal
channel to deliver useful energy. Applied Ocean Research, 26(5):198–204, 2004.
[5] IG Bryden, S Naik, P Fraenkel, and CR Bullen. Matching tidal current plants to
local flow conditions. Energy, 23(9):699–709, 1998.
[6] R Carballo, G Iglesias, and A Castro. Numerical model evaluation of tidal stream
energy resources in the ría de muros (nw spain). Renewable Energy, 34(6):1517–
1524, 2009.
[7] Michael Carter, Stephen P Bentley, et al. Correlations of soil properties. Pentech
press publishers, 1991.
[8] R Casciaro and L Cascini. Limit analysis by incremental-iterative procedure. In
IUTAM Int. Symp. on Granular Soils, Delft, 1982.
[9] Roger H Charlier. A “sleeper” awakes: tidal current power. Renewable and
Sustainable Energy Reviews, 7(6):515–529, 2003.
[10] RH Chen and J-L Chameau. Three-dimensional limit equilibrium analysis of
slopes. Geotechnique, 33(1):31–40, 1983.
[11] Wai-Fah Chen. Limit analysis and soil plasticity. Elsevier, 2013.
[12] YM Cheng, T Lansivaara, and WB Wei. Two-dimensional slope stability analysis

36
 by limit equilibrium and strength reduction methods. Computers and Geotechnics,
34(3):137–150, 2007.
[13] CA Coulomb. Essai sur une application des regles de maximis et minimis a
quelques problemes de statique relatifs a l’architecture (essay on maximums and
minimums of rules to some static problems relating to architecture). 1973.
[14] Ethan Dawson, Kwangho You, and Yeonjun Park. Strength-reduction stability
analysis of rock slopes using the hoek-brown failure criterion. Geotechnical special
publication, pages 65–77, 2000.
[15] R De Borst. Calculation of collapse loads using higher order elements. In IUTAM
Symposium on Deformation and Failure of Granular Materials, pages 503–513.
Balkema Rotterdam, 1982.
[16] R De Borst and PA Vermeer. Possibilites and limitations of finite elements for limit
analysis. Geotechnique, 34(2):199–210, 1984.
[17] Daniel Charles Drucker. Coulomb friction, plasticity, and limit loads. Technical
report, DTIC Document, 1953.
[18] James Michael Duncan. State of the art: limit equilibrium and finite-element
analysis of slopes. Journal of Geotechnical engineering, 122(7):577–596, 1996.
[19] PL Fraenkel. Marine current turbines: pioneering the development of marine
kinetic energy converters. Proceedings of the Institution of Mechanical Engineers,
Part A: Journal of Power and Energy, 221(2):159–169, 2007.
[20] DV Griffiths and PA Lane. Slope stability analysis by finite elements. Geotechnique,
49(3):387–403, 1999.
[21] Linus Hammar, Jimmy Ehnberg, Alberto Mavume, Boaventura C Cuamba, and
Sverker Molander. Renewable ocean energy in the western indian ocean. Renewable
and Sustainable Energy Reviews, 16(7):4938–4950, 2012.
[22] GT Houlsby. How the dilatancy of soils affects their behaviour. University of
Oxford, Department of Engineering Science, 1991.
[23] Guy T Houlsby, Lars Bo Ibsen, and Byron W Byrne. Suction caissons for wind
turbines. Frontiers in Offshore Geotechnics: ISFOG, Perth, WA, Australia, pages
75–93, 2005.
[24] Chul-Hee Jo, Do-Youb Kim, Kang-Hee Lee, Yu-Ho Rho, and Kook-Hyun Kim.
Design of horizontal axis tidal current power turbine with wake analysis. Journal
of the Korean society for New and Renewable Energy, 7(3):92–100, 2011.

37
[25] Chul-Hee Jo, Do-Youb Kim, Yu-Ho Rho, Kang-Hee Lee, and Cameron Johnstone.
Fsi analysis of deformation along offshore pile structure for tidal current power.
Renewable Energy, 54:248–252, 2013.
[26] MJ Khan, G Bhuyan, MT Iqbal, and JE Quaicoe. Hydrokinetic energy conversion
systems and assessment of horizontal and vertical axis turbines for river and tidal
applications: A technology status review. Applied Energy, 86(10):1823–1835,
2009.
[27] S Kiho, M Shiono, and K Suzuki. The power generation from tidal currents by
darrieus turbine. Renewable Energy, 9(1):1242–1245, 1996.
[28] J King and T Tryfonas. Tidal stream power technology-state of the art. In OCEANS
2009-EUROPE, pages 1–8. IEEE, 2009.
[29] RT Koekkoek. Gravity Base Foundations for Offshore Wind Turbines. PhD thesis,
TU Delft, Delft University of Technology, 2015.
[30] A. V. Lyamin and S. W. Sloan. Upper bound limit analysis using linear finite
elements and non-linear programming. International Journal for Numerical and
Analytical Methods in Geomechanics, 26(2):181–216, 2002.
[31] John Lysmer. Limit analysis of plane problems in soil mechanics. Journal of Soil
Mechanics & Foundations Div, 1970.
[32] David JC MacKay. Enhancing electrical supply by pumped storage in tidal lagoons.
Cavendish Laboratory, University of Cambridge, 2007.
[33] N Manoharan and SP Dasgupta. Bearing capacity of surface footings by finite
elements. Computers & Structures, 54(4):563–586, 1995.
[34] Joaquin Marti and Peter Cundall. Mixed discretization procedure for accurate
modelling of plastic collapse. International Journal for Numerical and Analytical
Methods in Geomechanics, 6(1):129–139, 1982.
[35] Allan Mason-Jones, Daphne Maria O’Doherty, Ceri E Morris, and Timothy
O’Doherty. Influence of a velocity profile and support structure on tidal stream
turbine performance. Renewable Energy, 52:23–30, 2013.
[36] GG Meyerhof. Penetration tests and bearing capacity of cohesionless soils. Journal
of the Soil Mechanics and Foundations Division, 82(1):1–19, 1956.
[37] JAC Orme and I Masters. Analysis and comparison of support structure concepts for
tidal stream turbines. In Proceedings of World Maritime Technology Conference,
2006.

38
[38] JAC Orme, I Masters, and C MATH. Design and testing of a direct drive tidal
stream generator. In Proceedings of the Institute of Marine Engineering, Science
and Technology. Part B, Journal of marine design and operations, number 9, pages
31–36. Institute of Marine Engineering, Science and Technology, 2005.
[39] Jim Platts. Fueling our future. an introduction to sustainable energy. Environmental
Conservation, 34(04):344–344, 2007.
[40] Junuthula Narasimha Reddy. An introduction to the finite element method, volume 2.
McGraw-Hill New York, 1993.
[41] Fergal O Rourke, Fergal Boyle, and Anthony Reynolds. Renewable energy re-
sources and technologies applicable to ireland. Renewable and Sustainable Energy
Reviews, 13(8):1975–1984, 2009.
[42] Marie Lunde Sæterstad. Dimensioning loads for a tidal turbine. 2011.
[43] M Sánchez, R Carballo, V Ramos, and G Iglesias. Energy production from tidal
currents in an estuary: A comparative study of floating and bottom-fixed turbines.
Energy, 77:802–811, 2014.
[44] Rüdiger Scharff and Michael Siems. Monopile foundations for offshore wind
turbines–solutions for greater water depths. Steel Construction, 6(1):47–53, 2013.
[45] Marc Senders. Tripods with suction caissons as foundations for offshore wind
turbines on sand. In Frontiers in Offshore Geotechnics: Proceedings of the Inter-
national Symposium on Frontiers in Offshore Geotechnics (IS-FOG 2005), 19-21
Sept 2005, Perth, WA, Australia, page 397. CRC Press, 2005.
[46] Daichao Sheng and Scott W Sloan. Load stepping schemes for critical state models.
International Journal for Numerical Methods in Engineering, 50(1):67–93, 2001.
[47] SW Sloan. Geotechnical stability analysis. Géotechnique, 63(7):531–571, 2013.
[48] SW Sloan and MF Randolph. Numerical prediction of collapse loads using finite
element methods. International Journal for Numerical and Analytical Methods in
Geomechanics, 6(1):47–76, 1982.
[49] Karl Terzaghi. Theoretical soil mechanics. Wiley, 1943.
[50] CT Toh and SW Sloan. Finite element analyses of isotropic and anisotropic cohesive
soils with a view to correctly predicting impending collapse. International Journal
for Numerical and Analytical Methods in Geomechanics, 4(1):1–23, 1980.
[51] Hai-Sui Yu. Plasticity and geotechnics, volume 13. Springer Science & Business
Media, 2007.

39