Zeidan Damchapter 2014

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/281320646
Design and Analysis of Concrete Gravity Dams
Research · August 2015

DOI: 10.13140/RG.2.1.4676.5289
CITATIONS READS
3 20,649
1 author:
Bakenaz A. Zeidan
Tanta University
120 PUBLICATIONS 153 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
The Nile River Book View project
Isotope hydrology and water resources View project
All content following this page was uploaded by Bakenaz A. Zeidan on 29 August 2015.
The user has requested enhancement of the downloaded file.

Dr. Bakenaz A. Zeidan –State of Art in Design and Analysis of Concrete Gravity Dams 2014
Article Review
State of Art
Design and Analysis of Concrete
Gravity Dams
By
Dr. Bakenaz A. Zeidan
2014
1
LIST OF CONTENTS
Summary
1. Introduction to Gravity Dams
2. About Concrete Gravity Dams
3. Cases of Loading on Gravity Dams
4. Theoretical Approach Gravity Dams
5. Modeling of Gravity Dam
6. Analysis of Gravity Dams
7. Safety Criteria for Gravity Dams
8. Recent Trends in Gravity Dams
9. Summing up
2
State of Art in Design and Analysis of

Concrete Gravity Dams
1. INTRODUCTION
Dams have been constructed for millennia, influencing the lives of humans and the
ecosystems they inhabit. Remnants of one such man-made structure dating back 5,000 years are still
standing in northeast Africa (UNESCO-WWAP, 2003). Around 2950-2750 B.C., the first dam
known to exist was built by the ancient Egyptians, measuring 11.3 meters [m] (37 feet [ft]) tall, with
a crest length of 106 m (348 ft) and foundation length of 80.7 m (265 ft) (Yang, et al, 1999). The
dam was composed of 100,000 tons of rubble, gravel, and stone, with an outer shell of limestone.
The immense weight was enough to contain water in a reservoir estimated to have been 570,000
cubic meters [m3] (20 million cubic feet [ft3] or 460 acre-feet) in capacity (Yang, et al, 1999).
Many concrete gravity dams have been in service for over 50 years, and over this period important
advances in the methodologies for evaluation of natural phenomena hazards have caused the design-
basis events for these dams to be revised upwards. Older existing dams may fail to meet revised
safety criteria and structural rehabilitation to meet such criteria may be costly and difficult. The
identified causes of failure, based on a study of over 1600 dams [1] are: foundation problems (40%),
inadequate spillway (23%), poor construction (12%), uneven settlement (10%), and high pore
pressure (5%), acts of war (3%), embankment slips (2%), defective materials (2%), incorrect
operation (2%), and earthquakes (1%). Earthquake is a natural disaster that has claimed so many
lives and destroyed lots of property. Earthquake hazards had caused the collapse and damage to
continual functioning of essential services such as communication and transportation facilities,
buildings, dams, electric installations, ports, pipelines, water and waste water systems, electric and
nuclear power plants with severe economic losses. However, the structural response of a material to
different loads determines how it will be economically utilized in the design process. This
necessitates the seismic analysis of concrete gravity dams. Finite element has been widely used in
seismic analysis of concrete gravity dams utilizing the most natural method based on the
Lagrangian–Eulerian formulation.
1.1. Implementation of Concrete Gravity Dams
1.2. Design Procedure of Concrete Gravity Dams
1.3. Safety Criteria of Concrete Gravity Dams
During the recent years, the seismic behaviour of concrete gravity dams was in the centre of
consideration of dam engineers. Numerous researches have been conducted in order to determine
how the dams behave against the seismic loads. Many achievements were obtained in the process of
analysis and design of concrete dams including dam-reservoir-foundation interaction during an
earthquake. The earthquake response of concrete gravity dam-reservoir-foundation system has been
addressed to study the effect of foundation flexibility and reservoir water body on the seismic
response of concrete gravity dams. Safety evaluation of dynamic response of dams is important for
3
most of researchers. When such system is

subjected to an earthquake,
hydrodynamic pressures are developed
on upstream face of the dam due to the
vibration of the dam and reservoir water.
Consequently, the prediction of the
dynamic response of dam to earthquake
loadings is a complicated problem and
depends on several factors, such as
interaction of the dam with rock
foundation and reservoir, the computer modelling and material properties used in the analysis.
Gravity dams are very important structures. The collapse of a gravity dam due to earthquake
ground motion may cause an extensive damage to property and life losses. Therefore, the proper
design of gravity dams is an important issue in dam engineering. An integral part of this procedure
is to accurately estimate the dam earthquake response. The prediction of the actual response of a
gravity dam subjected to earthquake is a very complicated problem. It depends on several factors
such as dam-foundation interaction, dam-water interaction, material model used and the analytical
model employed. In fluid-structure interaction one of the main problems is the identification of the
hydrodynamic pressure applied on the dam body during earthquake excitation. The analysis of dam-
reservoir system is complicated more than that of the dam itself due to the difference between the
characteristics of fluid and dam's concrete on one side and the interaction between reservoir and
dam on the other side.
The earthquake response of concrete gravity dam-reservoir-foundation system has been
addressed to study the effect of foundation flexibility and reservoir water body on the seismic
response of concrete gravity dams. Safety evaluation of dynamic response of dams is important for
most of researchers. When such system is subjected to an earthquake, hydrodynamic pressures are
developed on upstream face of the dam due to the vibration of the dam and reservoir water.
Consequently, the prediction of the dynamic response of dam to earthquake loadings is a
complicated problem and depends on several factors, such as interaction of the dam with rock
foundation and reservoir, the computer modeling and material properties used in the analysis.
For the structure on the rigid foundation, the input seismic acceleration gives rise to an overturning
moment and transverse base shear. As the rock is very stiff, these two stress resultants will not lead
to any (additional) deformation or rocking motion at the base. For the structure founded on flexible
soil, the motion of the base of the structure will be different from the free-field motion because of
the coupling of the structure-soil system. This process, in which the response of the soil influences
the motion of the structure and response of the structure influences the motion of the soil, is referred
to as soil-structure interaction (SSI) presented by Wolf (1985) [14]. The objective of this paper is to
assess the impact of foundation flexibility and dam-reservoir-foundation interaction on seismic
response of high concrete gravity dams.
4
2. ABOUT CONCRETE GRAVITY DAMS

There are two main types of dams, embankment dams and concrete dams. Embankment
dams can either be rock fill or earth fill dams. The method of construction is similar in both cases,
with just the main type of material differentiating. Concrete dams are superior in constructing
massive overflow discharge sections, and are therefore often used in areas where floods are
common. A lot less material is used compared to an embankment dam but concrete is usually more
costly. It is also easier to connect a hydropower station to a concrete dam. Three different height
spans exist when concrete dams are considered and they are defined as: low dams (up to 30 meters),
medium height dams (30-90 meters) and high dams (90 meters and above). This is a measurement
of the difference in elevation between the lowest constructed part of the dam foundation and the
walkway at the dam crest (1).
5
Figure 1: Different types of concrete dams (2).
A gravity dam is a solid structure, made of concrete or masonry, constructed across a river to
create a reservoir on its upstream. The section of the gravity dam is approximately triangular in
shape, with its apex at its top and maximum width at bottom. The section is so proportioned that it
resists the various forces acting on it by its own weight. Most of the gravity dams are solid, so that
no bending stress is introduced at any point and hence,
Figure they are sometimes
2: Concrete known
gravity dam north as
of solid gravity
Irkutsk, Russian
dams to distinguish them from hollow gravity dams in those hollow spaces are kept to reduce the
(www.hydroelecritc.energy.blogspot.com).
weight. Early gravity dams were built of masonry, but nowadays with improved methods of
construction, quality control and curing, concrete is most commonly used for the construction of
modern gravity dams. A gravity dam (Figure.1) is generally straight in plan and, therefore, it is also
called straight gravity dam. The upstream face is vertical or slightly inclined. The slope of the
downstream face usually varies between "0.7: 1‖ to ―0.8: 1‖. There are different types of concrete
dams based on the principal for the transfer of the hydrostatic pressure.
dams, Figure 2
Figure 3
Figure 4
6
The theory behind gravity dams is that their own weight should be sufficient to withstand
the hydrostatic pressures affecting them. This means that gravity dams are usually massive and
therefore require a lot of construction material. With the amount of concrete required, this dam type
may be somewhat expensive but on the other hand, it is very versatile. Another advantage is that it
can possess substantial overflow discharge capacity (1).
Buttress dams, shown in figure 8, are similar to gravity dams with the distinction that they also use
the gravity of the reservoir water instead of only the gravity of the dam itself. Because of this, the
dam body does not need to be as massive and use buttresses instead of a solid downstream part of
the dam. Being less solid on the downstream side, buttress dams have the advantage of being a lot
less affected by the water uplift force (1).
Arch dams, shown in figure 9, are curved around a vertical cord to resist the hydrostatic
pressure by arching thus transferring the pressure into the canyon walls. For this transfer to be
possible and cost effective the width to height ratio should not exceed 5:1, although in some cases
arch dams has been built with a ratio as high as 10:1. Another criteria which is important for arch
dams is the shape of the canyon, if it is symmetrical an arch dam is often very suitable. If the canyon
is a little less symmetrical, an arch dam with influences of a gravity dam may be constructed. If the
canyon is extremely asymmetric, another dam type may be preferred (1).
Figure 3: One of the buttresses in the Manic-Ceng buttress
2.1. Basic Definitions dam in Québec, Canada (www.dappolonia.com).
 Axis of the dam

The axis of the gravity dam is the line of the
upstream edge of the top (or crown) of the
dam. If the upstream face of the dam is
vertical, the axis of the dam coincides with
the plan of the upstream edge. In plan, the
axis of the dam indicates the horizontal trace
of the upstream edge of the top of the dam.
Figure 4: Arch dam in Zernez, Switzerland, view from side,
The axis of the dam in plan is also called the (www.commandatastorage.googleapsis.com).
base line of the dam. The axis of the dam in
plan is usually straight. However, in some
special cases, it may be slightly curved
upstream, or it may consist of a combination
of slightly curved right portions at ends and a central abutment straight portion to take the best
advantages of the topography of the site.
 Length of the dam
The length of the dam is the distance from one abutment to the other, measured along the axis of the
dam at the level of the top of the dam. It is the
usual practice to mark the distance from the left
abutment to the right abutment. The left
abutment is one which is to the left of the
person moving along with the current of water.
 Structural height of the dam:
The structural height of the dam is the
difference in elevations of the top of the dam
and the lowest point in the excavated
foundation. It, however, does not include the
depth of special geological features of
foundations such as narrow fault zones below
the foundation. In general, the height of the dam
7
means its structural height.

 Maximum base width of the dam:
The maximum base width of the dam is the maximum horizontal distance between the heel and the
toe of the maximum section of the dam in the middle of the valley.
 Toe and Heel:
The toe of the dam is the downstream edge of the base, and the heel is the upstream edge of the
base. When a person moves along with water current, his toe comes first and heel comes later.
 Hydraulic height of the dam
The hydraulic height of the dam is equal to the difference in elevations of the highest controlled
water surface on the upstream of the dam (i. e. FRL) and the lowest point in the river bed (3).
2.2. Dam Concrete Static Properties

2.2.1. Strength
A gravity dam should be constructed of concrete that will meet the design criteria for
strength, durability, permeability, and other required properties. Because of the sustained loading
generally associated with them, the concrete properties used for the analyses of static loading
conditions should include the effect of creep. Properties of concrete vary with age, the type of
cement, aggregates, and other ingredients as well as their proportions in the mix. Since different
concretes gain strength at different rates, measurements must be made of specimens of sufficient
age to permit evaluation of ultimate strengths. Although the concrete mix is usually designed for
only compressive strength, appropriate tests should be made to determine the tensile and shear
strength values (4).
2.2.2. Elastic Properties
Poisson’s ratio, the sustained modulus of elasticity of the concrete, and the latter’s ratio to
the deformation modulus of the foundation have significant effects on stress distribution in the
structure. Values of the modulus of elasticity, although not directly proportional to concrete
strength, do follow the same trend, with the higher strength concretes having a higher value for
modulus of elasticity. As with the strength properties, the elastic modulus is influenced by mix
proportions, cement, aggregate, admixtures, and age. The deformation that occurs immediately with
application of load depends on the instantaneous elastic modulus. The increase in deformation
which occurs over a period of time with a constant load is the result of creep or plastic flow in the
concrete. The effects of creep are generally accounted for by determining a sustained modulus of
elasticity of the concrete for use in the analyses of static loadings. Instantaneous moduli of elasticity
and Poisson’s ratios should be determined for the different ages of concrete when the cylinders are
initially loaded. The sustained modulus of elasticity should be determined from these cylinders after
specific periods of time under constant sustained load. These periods of loading are often 365 and
730 days. The cylinders to be tested should be of the same size and cured in the same manner as
those used for the compressive strength tests. The values of instantaneous modulus of elasticity,
Poisson’s ratio, and sustained modulus of elasticity used in the analyses should be the average of all
test cylinder values (4).
2.2.3. Thermal Properties
The effects of temperature change in gravity dams are not as important in the design as those
in arch dams. However, during construction, the temperature change of the concrete in the dam
should be controlled to avoid undesirable cracking. Thermal properties necessary for the evaluation
of temperature changes are the coefficient of thermal expansion, thermal conductivity, specific
heat, and diffusivity.The coefficient of thermal expansion is the length change per unit length for 1
degree temperature change. Thermal conductivity is the rate of heat conduction through a unit
8
thickness over a unit area of the material subjected to a unit temperature difference between faces.
The specific heat is defined as the amount of heat required to raise the temperature of a unit mass of
the material 1 degree. Diffusivity of concrete is an index of the facility with which concrete will
undergo temperature change. The diffusivity is calculated from the values of specific heat, thermal
conductivity, and density. Appropriate laboratory tests should be made of the design mix to
determine all concrete properties (4).
2.3. Dam Concrete Dynamic Properties (USBR)

2.3.1. Strength
No data are yet available to indicate what the strength characteristics are under dynamic
loading.
2.3.2. Elastic Properties
Until dynamic modulus information is available, the instantaneous modulus of elasticity
determined for concrete specimens at the time of initial loading should be the value used for
analyses of dynamic effects.
2.3.3. Average Properties
Necessary values of concrete properties may be estimated from published data for
preliminary studies until laboratory test data are available. Until long-term tests are made to
determine the effects of creep, the sustained modulus of elasticity should be taken as 60 to 70
percent of the laboratory value for the instantaneous modulus of elasticity. Criteria-If no tests or
published data are available, the following average values for concrete properties may be used for
preliminary designs until test data are available for better results (USBR):
 Compressive strength-3,000 to 5,000 Ibs/in2 (20.7 to 34.5 MPa)
 Tensile strength-5 to 6 percent of the compressive strength
 Shear strength: Cohesion-about 10 percent of the compressive strength
 Coefficient of internal friction- l.0
 Poisson’s ratio- 0.2
 Instantaneous modulus of elasticity- 5.0 x 106 lbs/in2 (34.5 GPa)
 Sustained modulus of elasticity- 3.0 x 106 lbs/in2 (20.7 GPa)
 Coefficient of thermal expansion- 5.0 x 10-6/“F (9.0 x l0-6PC)
 Unit weight- 150 Ibs/ft3 (2402.8 kg/m3)
2.4. Foundation Properties

2.4.1. Deformation Modulus
Foundation deformations caused by loads from the dam affect the stress distributions within
the dam. Conversely, response of the dam to external loading and foundation deformability
determines the stresses within the foundation. Proper evaluation of the dam and foundation
interaction requires as accurate a determination of foundation deformation characteristics as
possible. Although the dam is considered to be homogeneous, elastic, and isotropic, its foundation
is generally heterogeneous, inelastic, and anisotropic. These characteristics of the foundation have
significant effects on the deformation moduli of the foundation. The analysis of a gravity dam
should include the effective deformation modulus and its variation over the entire contact area of
the dam with the foundation. The deformation modulus is defined as the ratio of applied stress to
elastic strain plus inelastic strain and should be determined for each foundation material. The
effective deformation modulus is a composite of deformation moduli for all materials within a
particular segment of the foundation. Good compositional description of the zone tested for
9
deformation modulus and adequate geologic logging of the drill cores permit extrapolation of results
to untested zones of similar material.
Criteria- The following foundation data should be obtained for the analysis of a gravity dam (4):
 The deformation modulus of each type of material within the loaded area of the foundation.
 The effects of joints, shears, and faults obtained by direct (testing) or indirect (reduction
factor) methods.
 An effective deformation modulus, as determined when more than one type of material is
present in a foundation.
 The effective deformation moduli, as determined at enough locations along the foundation
contact to provide adequate definition of the variation in deformability and to permit
extrapolation to untested areas when necessary.
2.4.2. Shear Strength
Resistance to shear within the foundation and between the dam and its foundation depends
upon the cohesion and internal friction inherent in the foundation materials and in the bond between
concrete and rock at the contact with the dam. These properties are determined from laboratory and
in situ tests. The results of laboratory triaxial and direct shear tests, as well as in situ shear tests, are
generally reported in the form of the Coulomb equation:
R = C.A + N. tan φ or
(Shear resistance) = (unit cohesion times area) +
(effective normal force times coefficient of internal friction)
which defines a linear relationship between shear resistance and normal load. The value of shear
resistance obtained as above should be limited to use for the range of normal loads used for the
tests. Although this assumption of linearity is usually realistic for the shear resistance of intact rock
over the range of normal loads tested, a curve of shear resistance versus normal load should be used
for materials other than intact rock. The shear resistance versus normal load relationship is
determined from a number of tests at different normal loads. The individual tests give the
relationship of shear resistance to displacement for a particular normal load. The results of these
individual tests are used to obtain a shear resistance versus normal load curve. The displacement
used to determine the shear resistance is the maximum displacement that can be allowed on the
possible sliding plane without causing unacceptable stress concentrations within the dam. Since
specimens tested in the laboratory or in situ are small compared to the foundation, the scale effect
should be carefully considered in determining the values of shear resistance to be used.
When a foundation is nonhomogeneous, the possible sliding surface may consist of several
different materials, some intact and some fractured. Intact rock reaches its maximum break bond
resistance with less deformation than is necessary for fractured materials to develop their maximum
frictional resistances. Therefore, the shear resistance developed by each fractured material depends
upon the displacement of the intact rock part of the surface. If the intact rock shears, the shear
resistance of the entire plane is equal to the combined sliding frictional resistance for all materials
along the plane (4).
2.4.3. Pore Pressure and Permeability

Analysis of a dam foundation requires a knowledge of the hydrostatic pressure distribution
in the foundation. Permeability is controlled by the characteristics of the rock type, the jointing
systems, the shears and fissures, fault zones, and, at some dam sites, by solution cavities in the rock.
The exit gradient for shear zone materials that surface near the downstream toe of the dam should
also be determined to check against the possibility of piping. Laboratory values for permeability of
sample specimens are applicable only to the portion or portions of the foundation that they
10
represent. Permeability of the aforementioned geologic features can best be determined by in situ
testing. The permeability obtained are used in the determination of pore pressures for analyses of
stresses, stability, and piping. Such a determination may be made by several methods including two-
and three dimensional physical models, two- and three-dimensional finite element models, and
electric analogs. If foundation grouting and drainage or other treatment are to be used, their effects
on the pore pressures should be included (4).
2.4.4. Treatment
Foundation treatment is used to correct deficiencies and improve physical properties by
grouting, drainage, excavation of inadequate materials, reinforcement and backfill with concrete.
Some reasons for foundation treatment are: (1) improvement of deformation moduli, (2) prevention
of sliding of foundation blocks, (3) prevention of relative displacement of foundation blocks, (4)
prevention of piping and reduction of pore pressures, and (5) provision of an artificial foundation in
the absence of adequate materials. Regardless of the reason for the foundation treatment, its effects
on the other foundation properties should be considered in the analyses (4).
2.4.5. Compressive and Tensile Strength

Compressive strength of the foundation rock can be an important factor in determining
thickness requirements for a dam at its contact with the foundation. Where the foundation rock is
nonhomogeneous, tests to obtain compressive strength values should be made for each type of rock
in the loaded portion of the foundation. A determination of tensile strength of the rock is seldom
required because unhealed joints, shears, etc., cannot transmit tensile stress within the foundation.
3. GRAVITY DAM LOAD COMBINATION

Factors to be considered as contributing to the loading combinations for a gravity dam are:
(1) reservoir and tail water loads, (2) temperature, (3) internal hydrostatic pressure, (4) dead weight,
(5) ice, (6) silt, and (7) earthquake. Such factors as dead weight and static water loads can be
calculated accurately. Others such as earthquake, temperature, ice, silt, and internal hydrostatic
pressure must be predicted on the basis of assumptions of varying reliability (USBR).
3.1. Static Loads:

There are several types of forces acting on dams, the eight that are commonly used are listed
below. In figure (5) the forces acting on a gravity dam are shown (1):
 Dead weight
Figure (5)
 Hydrostatic pressure from reservoir
 Hydrostatic pressure from tail water
 Internal hydrostatic pressure (uplift pressure)
 Sand and silt
 Ice
 Temperature
 Earthquake
11
Dead weight is the gravity affecting the dam itself, since we are dealing with compact, heavy
structures this is a major factor. This is a static load as well as the others, except for earthquake.
Hydrostatic pressure from reservoir is the
pressure created by the upstream water.
Hydrostatic pressure from tail water affects the
dam the same way as the hydrostatic pressure
caused by reservoir water does, with the only
difference that it is the pressure from the water
downstream of the dam.
Internal hydrostatic pressure is what we also
call uplift pressure. This is the pressure from the
water in the foundation and in the dam body that
will push the dam body upward, causing an
enhanced risk of sliding. This is especially
problematic for gravity dams since they cover a
greater area than other types of concrete dams,
Figure (6).
Sand and silt is the load applied as an earth
pressure from eroded material at the upstream face
of the dam. It is usually reasonably small
compared to the hydrostatic reservoir pressure.
Ice load affects the upstream face of the dam, as the surface of the water freezes. If the ice gets thick
enough it will cause a significant load. load is concentrated to a small surface where the dam body is
thinnest.
Temperature is a concern both during the construction phase and the entire lifespan of the dam.
The hardening of the concrete causes severe temperature variations that will lead to strains.
Depending on where in the world the dam is located the temperature changes may continue, during
the entire lifespan of the dam, to be an important issue.
Earthquake is a dynamic load. This load is, unlike the other seven, very hard to predict but is very
important to consider in earthquake affected regions.
3.2. Seismic Loads
Figure (7)
12
Earthquake or seismic loads are the major dynamic loads (4) being considered in the analysis
and design of dams especially in earthquake prone areas. The seismic coefficient method is used in
determining the resultant location and sliding stability of dams. Seismic analysis of dams is
performed for the most unfavourable direction, despite the fact that earthquake acceleration might
take place in any direction. Figure (7) shows the seismic coefficient α for dynamic loads on a
gravity dam. There are different ways of computing earthquake loads on dams. The deterministic
approach may be employed where the ground acceleration in terms of g (acceleration due to gravity)
is specified for the region where the dam will be constructed. Hence, the excitingFigure (8) on the
force
structure is (3), Figure (8),
P(t) = Max
(1)
and
ax = αg
(2)
where ax, α, g are the ground

acceleration, seismic coefficient
and acceleration due to
gravity respectively. From
Fig.10, the equilibrium system
is expressed as:
Pex=Max=Wαg/g=Wα (3a)
The hydrodynamic pressure exerted along dam-reservoir interface is given by:
Pew =(2 * Ce * α * y * √ (h *y)) / 3 (4a)
where
Ce = 51 / √ (1 – 0.72 * (h / (1000te))2) (4b)
where Pex, M, ax, W, α, g are the horizontal earthquake force on the dam, mass horizontal
earthquake acceleration, weight, acceleration due to gravity and seismic coefficient respectively.
Also Pew, h, te are the additional total water load down to depth y, total height of reservoir, and
period of vibration respectively (3).
3.3. Load Combinations

Gravity dams should be designed for all appropriate load combinations, using the proper
safety factor for each. Combinations of transitory loads, each of which has only a remote probability
of occurrence at any given time, have negligible probability of simultaneous occurrence and should
not be considered as an appropriate load combination. Temperature loadings should be included
when applicable, as previously discussed in this monograph, Figure (9).
(1) Usual loading combinations.-Normal design reservoir elevation with appropriate dead loads,
uplift, silt, ice, and tail water. If temperature loads are applicable, use minimum usual temperatures
occurring at that time.
(2) Unusual loading combinations .-Maximum design reservoir elevation with appropriate dead
loads, silt, tail water, uplift, and minimum usual temperatures occurring at that time, if applicable.
(3) Extreme loading combinations.-The usual loading plus effects of the ―Maximum Credible
Earthquake.‖
13
(4) Other loadings and investigations.-(a) The usual or unusual loading combination with drains
inoperative. (b) Dead load. (c) Any other
loading combination which, in the
designer’s opinion, should be analyzed for
a particular dam.
From the loads mentioned above it is
possible to create load cases as:
Figure (9)
A usual load case occurs often, or

even all the time. For example the
combination of dead weight,
sand and silt load, uplift pressure,
hydraulic pressures from reservoir
and tail water at a normal level.
An unusual load case may be the
usual case from above with added
ice load and lowest possible
temperature.
An extreme load case could be a
combination of the worst scenario in all eight load-types, including a nearby earthquake. The reason
to use these load cases is to be able to estimate and calculate safety factors, the more usual a load
case is, the higher the safety factor should be. This is just an example of how different scenarios can
be predicted, in reality these different cases are very thoroughly evaluated, with a lot of different
combinations (1).
3.4. Stability criteria

The loads listed in section 2.1 will create different types of stresses in the dam body.
Although every dam project is unique, problems with these stresses will often occur in the same
areas. Figure 10 shows these general critical areas. To create a clear overview of the figure none of
14
the applied loads are displayed (1). To evaluate shear stress in the different, carefully chosen, areas
in and below the dam foundation, information about the friction coefficient both in concrete, rock
and between the two is needed. The cohesion in concrete and rock is also needed. With the vertical
stress, the shear stress, the friction coefficient and the cohesion a safety factor, K can be calculated.
This safety factor is calculated
according to Mohr-coulomb failure criterion.
Unstable sliding surfaces can occur in numerous places in the dam and the foundation. Therefore it
is important to single out the areas where the greatest risk of damage exists. For example such
surfaces could be cracks in the ground, where a change of rock material occurs and in various places
in the dam, the two most obvious of such being in the foundation plane just in the contact surface
with the underlying rock and the horizontal plane where the slope of the downstream side of the
dam body starts to flatten out shown in figure 10 (1).
Stability criteria for concrete gravity dams accounts for four types of controls to be
considered as follows:
Sliding stability; To make sure the calculations are accurate, the element standards, described later
in this chapter, will be considered when creating elements inside and around the critical areas,
especially close to the dam foundation and in the batter (5).
Tension stress often occurs, in the region around the dam heel. When analyzing the tension stress
we use a simplification for FEM modeling that states that the number of elements with tension in
the bottom layer of the dam cannot exceed seven per cent of the total amount of elements in that
layer. The reason that we can use this simplification is that the elements in the foundation have
roughly the same size, which leads to that the percentage of tension elements is considered the same
as the percentage.
Compression stresses are handled by looking at the whole model and then determine where the
greatest risk for compressive failure appears. Material characteristics have to be evaluated and
compared to the computed compressive stresses (5).
Displacement control is based on the entire model. Since for example the displacement in the top
of the dam depends on the displacement in the bottom part of the dam there is not really one area to
focus on to receive good displacement results.
3.5. Factors of Safety

All loads to be used in design should be chosen to represent, as nearly as can be determined,
the actual loads that will occur on the structure during operation, in accordance with the criteria
under ―Load Combinations.‖ Methods of determining load-resisting capacity of the dam should be
the most accurate available. All uncertainties regarding loads or load-carrying capacity should be-
resolved as far as practicable by field or laboratory tests and by thorough exploration and inspection
of the foundation. Thus, the factor of safety should be as accurate an evaluation as possible of the
capacity of the structure to resist applied loads. All safety factors listed are minimum values. Like
other important structures, dams should be regularly and frequently inspected. Adequate
observations and measurements should be made of the structural behavior of the dam and its
foundation to assure that the structure is functioning as designed.
Although somewhat lower safety factors may be permitted for limited local areas within the
foundation, overall safety factors for the dam and its foundation after beneficiation should meet
requirements for the loading combination being analyzed. For other loading combinations where
safety factors are not specified, the designer is responsible for selection of safety factors consistent
15
with those for loading combination categories previously discussed. Somewhat higher safety factors
should be used for foundation studies because of the greater amount of uncertainty involved in
assessing foundation load-resisting capacity. Safety factors for gravity dams are based on the use of
the gravity method of analysis and those for foundation sliding stability are based on an assumption
of uniform stress distribution on the plane being analyzed.
Criteria (USBR) Figure (11)

(I) Compressive stress.-The maximum allowable compressive stress for concrete in a gravity dam
subjected to any of the ―Usual Loading Combinations‖ should not be greater than the specified
compressive strength divided by a safety factor of 3.0. Under no circumstance should the allowable
compressive stress for the ―Usual Loading Combinations‖ exceed 1,500 lbs/in2 (10.3 MPa). A
safety factor of 2.0 should be used in determining the allowable compressive stress for the ―Unusual
Loading Combinations.‖ The maximum allowable compressive stress for the ―Unusual Loading
Combinations‖ should in no case exceed 2,250 lbs/in2 (15.5 MPa). The maximum allowable
compressive stress for the ―Extreme Loading Combinations‖ should be determined in the same way
using a safety factor greater than 1.0. Safety factors of 4.0, 2.7, and 1.3 should be used in
determining allowable compressive stresses in the foundation for ―Usual,‖ ―Unusual,‖ and
―Extreme Loading Combinations,‖ respectively, Figure (11).
(2) Tensile stress.-In order not to exceed the allowable tensile stress, the minimum allowable
compressive stress computed without internal hydrostatic pressure should be determined from the
following
expression which takes into account the tensile strength of the concrete at the lift surfaces:
σz = p. γ. h – (ft/s)
where:
σz = minimum allowable stress at the face
p = a reduction factor to account for drains
γ = unit weight of water
h = depth below water surface
ft = tensile strength of concrete at lift surfaces
s = safety factor.
All parameters must be specified using consistent units. The value of p should be 1.0 if
drains are not present or if cracking occurs at the downstream face and 0.4 if drains are used. A
16
safety factor of 3.0 should be used for ―Usual‖ and 2.0 for ―Unusual Loading Combinations.‖ The
allowable value of 0 σz for ―Usual Loading Combinations‖ should never be less than 0. Cracking
should be assumed to occur if the stress at the upstream face is less than σz, computed from the
above equation with a safety factor of 1.0 for the ―Extreme Loading Combinations.‖ The structure
should be deemed safe for this loading if, after cracking has been included, stresses in the structure
do not exceed the specified strengths and sliding stability is maintained.
4. THEORETICAL APPROACH
In both the Eulerian and Lagrangian methods, the governing fluid-structure system equation
is solved using wave propagation through the fluid by assuming linear incompressibility and
inviscousity (6), Figure (12).
4.1. Governing Equations
Assuming that water is linearly compressible and neglecting its viscosity, the small
amplitudeFigure (12)
irrotational motion of the water is governed by the two-dimensional wave equation
(7),(10):
Ω (1)
where is the acoustic hydrodynamic pressure; t is time and is the two-dimensional Laplace
operator and C is the speed of pressure wave given by:
4.2. Dam-Reservoir Boundary Condition:

In the common boundary between the reservoir and the dam body, an interaction between
these two boundaries occurs which is the result of an inertia force caused by the movement of the
reservoir wall. At the surface of fluid-structure, there must be no flow across the interface. This is
based on the fact that face of concrete dams is impermeable. Hence, the applied pressure on the
reservoir wall caused by the inertial force is as follow:
(3)
17
where is normal acceleration of the dam body on the upstream face and n is normal vector on the
interface of the dam-reservoir outwards the dam body and is the mass density of the reservoir
water.
4.3. Reservoir-Foundation Boundary Condition:

According to the rigidity of the reservoir bottom, by assuming the horizontal movement of
the earth, the pressure gradient is neglected. Reservoir bottom absorption effect is implemented as:
(4)
where is the damping coefficient characterizing the effects of absorption of hydrodynamic

pressure waves at the reservoir boundary and is the wave reflection coefficient, which represents
the ratio of the amplitude of the reflected wave to that of the normally incident pressure wave at the
reservoir is related to by the following expressions:
It is believed that a value from 1 to 0 would cover the wide range of materials encountered at
the boundary of actual reservoirs. The value of the wave reflection coefficient that characterizes
the reservoir bottom materials should be selected based on their actual properties, not on properties
of the foundation rock. Materials on the reservoir bottom has great influence in absorbing of
earthquake waves and decreases the system response under the vertical component of the earthquake
and this effect is also important for horizontal component.
4.4. Reservoir-Far-End Boundary Condition

With the vibration of the dam, volumetric hydrodynamic pressure waves are created in the
reservoir and propagate toward the upstream. If the length of the dam is assumed to be infinity, then
these waves would approach to vanish. It should be noted that the length of reservoir is assumed as
a finite length, in numerical modeling. Hence, an artificial boundary is applied to simulate effect of
an infinite reservoir. For modeling far-end truncated boundary, viscous boundary condition (called
as Sommerfeld boundary condition) is utilized to absorb completely the outgoing pressure waves
given as Somerfield-type radiation boundary condition may be implemented namely:
(5)
4.5. Free-Surface Boundary Condition

In high dams, surface waves are negligible and hydrodynamic pressure on the free surface is
set to be zero, the boundary condition is easily defined as:
P(x, y, z, t) = 0 (6)
The dam-reservoir interaction is solved by coupled solution procedure while the boundary
condition is applied at the reservoir’s far-end truncated boundary. The foundation is defined as a
different part from the structure with different modulus of elasticity. An efficient coupling
procedure is formulated by using the coupling coincide nodes method. Summerfield’s boundary
condition at the far end of the infinite fluid domain is implemented. Figure 13 shows the coupled
dam-reservoir- foundation problem idealization.
18
Figure13: Dam-Reservoir-Foundation System
5. FINITE ELEMENT MODELING

FEM modeling of concrete gravity dams is a method with a lot of advantages compared to
traditional structural dynamics and scale modeling. Compared to scale modeling the time and cost
issue is the main factor, it is a lot cheaper to construct a virtual model than a physical one. Also the
convenience of computer based models compared to the location and rarity of scale models provide
a significant advantage. Compared to structural mechanics FEM has a big advantage in the
alteration of both construction and external loads. Once a dam has been modeled in FEM it is
possible to experiment and change details about it without the need to restart the whole process.
This is still just an analysis of a single section in a static state of a dam; a lot of aspect is
because of that limitation not dealt with at all. Examples of these aspects are: discharge capacity,
temperature changes, cracks, earthquakes and fatigue of the concrete. FEM means Finite Element
Method and it is a way of turning real life objects, such as a dam construction, to a computable
model. In the FEM the object is divided into smaller elements which are calculated separately,
preferably by a computer. It is the density and shapes of these elements that determines the accuracy
of the FEM-model. From advanced mathematical models to simple models made of for example
clay, they are still just models. Models can be more or less accurate, but they will never behave
exactly as reality would, (2).
The size and shape of elements is utterly important, and some basic standards have been set
up to make it easier to create well-functioning elements. The two most common types of two-
dimensional elements are quad and tri elements. Tri elements are made from three different nodes
and contain only one integration point while quads, as implied by their name, are made of four
nodes and contain four integration points. Therefore quad elements are more accurate and are to be
preferred. In our models we use only quad elements as displayed in Figure 14. The height-width
relation should not exceed three to one, for the quad elements, and no interior angle should be less
than 45 degrees (7). In practice the limit may be reduced to 30 degrees and the result will still be
acceptable. According to the two shape standards above, the most precise element has the height-
width ratio of one to one and all interior angles 90 degrees, the perfect square. Usually, fitting all
elements into these standards is impossible and not even that important, a small percentage of the
19
elements may remain distorted. The element standards in critical regions of the model where the
mesh may also possibly be denser. Minor alteration in the model’s geometry can be made, in such a
way that they generate no significant difference in the results. These alterations should be done so
the elements could form a better shape, given height-width relation and their interior angle (2).
Figure 14: A screenshot from a typical dam model displayed in ANSYS (8).
5.1. Strain and Stress in FEM

The elements used in FEM processes can either be plane stress or plane strain elements. In
both plane stress and plane strain analyses there are three components that need consideration in the
xy-plane. Two of these components are normal stresses, one horizontal and one vertical, the last of
the components is shear stress. However there are differences between the stress and strain. In plane
stress analysis all components of stress, except the three mentioned, are zero which leads to no
addition to the internal work in the elements. In the case of plane strain analysis, the stress
orthogonal the xy-plane can vary. However because of the definition of plane strain, the strain in
that perpendicular direction to the plane does not exist. Given this correlation of these contributions
to the internal work in the elements will not be affected. This makes it comparable to plane stress.
benefit of this is that the xy-plane can be evaluated with the three main stress components at the end
of all computations (7).
5.2. FEM Formulation

Referring to the total hydrodynamic pressure during the earthquake against the upstream
face, it has been shown that during the initial earthquake phases, the hydrodynamic pressure is
higher at the upper part of the dam because of the prevailing effect of water compressibility. If the
dominating period of earthquake is long, the increase of the hydrodynamic pressure is negligible.
Under the same condition, however, earthquake can also generate overall oscillation of the fluid
mass, because of the inertia forces developed in the fluid body (7). In the present study, the standard
Finite Element technique is adopted utilizing Galerkin’s method in which the structure displacement
vector is discretized as
u= Nu (6a)
and the fluid is similarly discretized as
p= Np (6b)
20
where and are the nodal parameters of each field and Nu and Np are appropriate shape functions.
The discrete equations of the structure dynamic response following Galerkin method reads (7):
M + C +K - Q + f =0 (7)
In which M, C, K and f refer to mass matrix, damping matrix, stiffness matrix of the structure and
prescribed force vector respectively, where , and are displacement, velocity and acceleration
vectors respectively. The coupling term in Equation 7 arises due to the pressures specified on the
boundary reads [7]
= (8)
Matrix Q shown in equation (8) transforms the accelerations of the structure to fluid pressure and
also transforms the hydrodynamic pressure into applied loads on the structure to simulate fluid
structure interaction. In Equation 8 is the direction vector of the normal to the interface. Standard
Galerkin’s discretization applied to the fluid Equation (1) and its boundary conditions leads to [7]
S + + H + QT + q = 0 (9)
in which S, H and q are pseudo fluid mass matrix, pseudo fluid damping matrix, pseudo fluid
stiffness matrix and prescribed flux vector respectively which are given by
S=- (10a)
= (10a)
H= (10a)
where Q is identical to that of Equation 8 and and are nodal pressure vector, the first and
second order derivatives of nodal pressure vector with respect to time, respectively. Hence, the
coupled equation of the fluid-structure system based on Equations (7) and (9) subjected to
earthquake ground motion can be presented as follows (8):
+ (11)
In which represents the nodal ground acceleration vector.
5.3. Fluid-Structure Interaction

Earthquake-induced hydrodynamic pressures on upstream face of a dam are important
factors in design consideration. Assuming that the fluid is incompressible, Westergaard [9] was the
first who derived an analytical solution for the hydrodynamic pressure acting on a rigid dam with a
vertical upstream face as a result of horizontal harmonic ground motion. In this method for the
analysis of concrete dams, fluid is treated as an added mass to the body of the dam. In the last fifty
years, many researchers have extended the Westergaard’s (9) classical work to include more
physical parameters such as the compressibility of the fluid in the reservoir, the flexibility of the
21
dam, and reservoir bottom absorption. However, analytical solutions are rare and available only for
a reservoir with a simple geometry and boundary conditions. Numerical models proved to be
powerful techniques that can be used for dam-reservoir interaction problems with complex two and
three dimensional geometries as well as complicated boundary and initial conditions. Studies of
Chopra [10]-[11] showed that fluid incompressibility assumption does not predict correctly the
applied hydrodynamic pressure on the dam body. The early studies on 2-dimentional gravity dam
roots back to late 1970s, in which interaction effects were considered through the exact and non-
numerical solutions of the governing equations, Chopra et al. [12]-[13]. Zienkiewicz and Taylor
[14] solved the coupled governing fluid-structure equations utilizing the finite element method. For
modeling the upstream boundary of the reservoir, they used radiative boundaries of thermal
analysis. AKKÖSE [15] presented the linear and nonlinear responses of a selected arch subjected
to earthquake ground motion. The hydrodynamic effects on the dynamic response of arch dams
were investigated using step-by-step integration by the Lagrangian approach. Aznarez et al. [16]
studied the effects of reservoir bottom absorbent materials, on the dynamic analysis of fluid-
structure interaction problem through the use of the boundary element method in the frequency
domain. Du et al. [17] studied the nonlinear seismic response analysis of a foundation-arch dam
system. They found that the maximum dynamic response was obtained from their method is lower
than that of the common methods. Akkose et al. [18] have studied the effect of sloshing on the
nonlinear dynamic response of the arch dam. Seghir et al. [19] used the coupled finite element and
symmetric boundary element to model the interaction problem. Bonnet et al. [20] used the
combination of the finite element and boundary element methods to simulate the dam-reservoir
interaction in the frequency domain. They considered an elastic material behavior and a rigid
reservoir bottom. The results of the analysis showed good agreement between theoretical and
numerical methods. Shariatmadar et al. [21] studied hydrodynamic pressures induced due to seismic
forces and Fluid-Structure Interaction. The interaction of reservoir water-dam structure and
foundation bed rock are modeled using the ANSYS code. The analytical results obtained from over
twenty 2D finite element modal analysis of concrete gravity dam show that the accurate modeling of
dam-reservoir-foundation and their interaction considerably affects the modal periods, mode shapes
and modal hydrodynamic pressure distribution. Seleemah and co-authors, [22] studied the seismic
response of base isolated liquid storage ground and elevated tanks employing coupled fluid-
structure system via ANSYS code and a good agreement was attained between analytical and
numerical results. Akhaveissy et al. [23] studied the linear dynamic behavior of the Pine Flat
concrete dam in the time domain to analyze reservoir–dam interaction. Also the effect of fluid
compressibility and the Sommerfeld boundary are used to determine the hydrodynamic pressure.
Hamidian et al. [24] studied dam-water-foundation rock interaction effects on linear and nonlinear
earthquake response of arch dams system subjected to earthquake ground motion using the finite
element method involving the materially and geometrically nonlinear effects. Zeidan [25] studied
the seismic dam-reservoir interaction of concrete gravity dams using ANSYS software with
coupling coincide nodes on the interface, Fig. (14).
y LC
ГF
HC
Reservoir
Domain Ω
ГR HB
HF ГI
Dam
domain
ГB heel toe
R
x
22
LF LB
Figure 14: Dam-Reservoir coupled system (25).
5.4. Fluid-Structure-Foundation Interaction

Studies that present a method called Soil Structure Interaction Method (SSI) for estimation
of free field earthquake motions at the site of dams. This method neglects the presence of structure
(Dam) during the earthquake and assumes that the relative displacement at the truncated boundary is
zero, and shows that under these circumstances, the foundation just bears the inertia force and does
not bear the earthquake force. Recent studies on dam-reservoir-foundation interaction deal with the
boundary conditions of the foundation
and foundation interaction with both
0.10
reservoir and dam bodies. However,
the problem of fluid-soil-structure crest
interaction is a complicated problem dam toe
0.08
dam heel
and need more interest from
Max. hz. displacement m
researchers. More efficient methods

0.06
are required to properly assess the
safety of concrete gravity dams
located in regions with significant 0.04
seismicity (8). The methods used for
the analysis of concrete dams under
earthquake loading range from the 0.02
simple pseudo-static method initially

proposed by Westergaard (1933) to
0.00
advanced numerical methods that 0.5 5 50 500
include the well-known FEM. Ef/ Ec ratio
Westergaard [9] introduced an Figure 15
approach to determine approximately
the linear response of the dam-reservoir system by a number of masses that are added to the dam
body. The method proposed by Westergaard assumes that the hydrodynamic effect on a rigid dam is
equivalent to the inertial force resulting from a mass distribution added on the dam body. The dam-
reservoir system can be categorized as a coupled field system in a way that these two physical
domains interact only at their interface [26]. To simplify and economize the finite element modeling
of an infinite reservoir, the far-end boundary of the reservoir has to be truncated. Sommerfeld
boundary condition [27] is an appropriate boundary condition for the truncated part of the reservoir.
Hydrodynamic pressure in seismic response of dam-reservoir interaction in time domain has been
investigated [28]. Preliminary design and evaluation of concrete gravity sections is usually
performed using the simplified response spectrum method proposed by Fenves and Chopra [10-13].
A standard fundamental mode of vibration, representative of typical sections, is used in this method.
This mode shape does not take into account the foundation flexibility since it is representative of a
standard concrete gravity section on rigid foundation.
As an alternative, the first mode of vibration of the concrete section could be estimated
using a finite element model with massless foundation. Fenves and Chopra [26], [27] studied the
dam-reservoir-foundation rock interaction in a frequency domain linear analysis. In the work
presented by Gaun et. al [28], an efficient numerical procedure has been described to study the
dynamic response of a reservoir-dam-foundation system directly in the time domain. Ghaemian et.
al [29] showed that the effects of foundation’s shape and mass on the linear response of arch dams
23
are considerable. The dam–foundation interaction effects are typically presented by a ―standard‖
mass-less foundation model [30]. In this case, it is assumed that the displacement at the bottom of
the foundation vanishes and roller supports is placed at the vertical sides of the foundation. The
most widely used model for soil radiation damping is the one of Lysmer and Kuhlemeyer [31]. In
this model the foundation is wrapped by dashpots tuned to absorb the S and P waves. In this model,
modeling the radiation damping on the far–end boundary of the massed foundation, 2- node
elements as boundary elements are used to apply the lumped dashpot on the far–end nodes of the
massed foundation model. The viscous boundary condition is applied on the far–end boundary of
the foundation to prevent the wave reflection form the artificial boundary of the infinite media in
finite element analysis (32). The most common soil–structure interaction (SSI) approach is based on
the ―added motion‖ formulation. This formulation is valid for free–field motions caused by
earthquake waves generated from all sources. The method requires that the free–field motions at the
base of the structure be calculated prior to the soil–structure interaction analysis [33]. Zeidan [34]
studied the seismic dam-reservoir-foundation interaction of concrete gravity dams using ANSYS
software with different ratios for Ec/Ef representing foundation flexibility, Fig. (15).
5.5. FEM Modeling Assumptions

In FEM simulation, in order to satisfy the continuity conditions between the fluid and solid
media at the boundaries. The nodes at the common lines of the fluid and the plane elements are
constrained to be coupled in the direction normal to the interface. Relative movements are allowed
to occur in the tangential directions. This is implemented by attaching the coincident nodes at the
common lines of the fluid and the plane elements in the normal direction. At the interface of the
fluid-structure system, only the displacements in the direction normal to the interface are assumed to
be compatible in the structure as well as the fluid. The fluid is generally assumed to be linear-
elastic, incompressible, irrotational and nonviscous. 2-D finite element model is implemented.
Absorption is considered at reservoir bottom. Since the extent of the reservoir is large, it is
necessary to truncate the reservoir at a sufficiently large distance from the dam. A length of
reservoir equivalent to two to three times its depth is chosen for adequate representation of
hydrodynamic effects on the dam body (7). The depth of foundation is taken as 1.5 the dam base
width into account in the calculations. The dam and foundation materials are assumed to be linear-
elastic, homogeneous and isotropic. The effect of foundation flexibility is considered for dam-
foundation rock interaction ratios i.e. modulus of elasticity of foundation to modulus of elasticity of
dam Ef/Ec.
6. ANALYSIS OF CONRETE GRAVITY DAMS

Selection of the method of analysis should be governed by the type and configuration of the
structure being considered. The gravity method will generally be sufficient for the analysis of most
structures, however, more sophisticated methods may be required for structures that are curved in
plan, or structures with unusual configurations.
5.1. Gravity Method

The gravity method assumes that the dam is a 2 dimensional rigid block. The foundation
pressure distribution is assumed to be linear. It is usually prudent to perform gravity analysis before
doing more rigorous studies. In most cases, if gravity analysis indicates that the dam is stable, no
further analyses need be done. A Stability criteria and required factors of safety for sliding are
required.
24
5.2. Finite Element Method

In most cases, the gravity analysis
method discussed above will be sufficient
for the determination of stability.
However, dams with irregular geometries
or spillway sections with long aprons may
require more rigorous analysis. The Finite Figure (16)
Element Method (FEM) permits the
engineer to closely model the actual
geometry of the structure and account for
its interaction with the foundation. For
example, consider the dam in Figure 16.
Note that the thinning spillway that forms
the toe of the dam is not stiff enough to
produce the foundation stress distribution
assumed in the gravity method. In this case, gravity analysis alone would have under-predicted base
cracking.
Finite element analysis allows not only modeling of the dam, but also the foundation rock
below the dam. One of the most important parameters in dam/foundation interaction is the ratio of
the modulus of deformation of the rock to the modulus of elasticity of the dam concrete. Figure 17
illustrates the effect that this ratio has on predicted crack length. As the modular ratio varies, the
amount of predicted base cracking varies also. As can be seen in Figure 17, assuming a low
deformation modulus (Er), is not necessarily conservative.
However, it is implicitly assumed that shear stress is distributed uniformly across the base.
This assumption is arbitrary and not very accurate. Finite element modeling can give some insight
into the distribution of base contact stress. As can be seen in figure 17, shear stress is at a maximum
at the tip of the propagating base crack. In this area, normal stress is zero, thus all shear resistance
must come from cohesion. Also, the peak shear stress is about twice the average shear stress. An un-
zipping failure mode can be seen here, as local shear strength is exceeded near the crack tip, the
crack propagates causing shear stress to increase in the area still in contact.
6.3. Modal Analysis and Natural Response

The structural response of a material to different loads determines how it will be economically
utilized in the design process. Earthquake is a natural disaster that has claimed so many lives and
Figure (17)
destroyed lots of property. Earthquake hazards had caused
the collapse and damage to continual functioning of
essential services such as communication and
transportation facilities, buildings, dams, electric
installations, ports, pipelines, water and waste water
systems, electric and nuclear power plants with severe
economic losses. Earthquake is a major source of seismic
forces that impinge on structures others are Tsunami,
seethe etc. Earth wall is chosen as a material for the dam
since its major constituent earth is abundantly available and
provides a sustainable solution. This necessitates the
seismic analysis of concrete gravity dam. Earthquakes had
caused severe damages and consequently huge economic
losses including losses of lives (3). Figure (18) shows mode
25
shapes for a gravity dam with empty reservoir while Figure (19) shows mode shapes for a gravity dam with
full reservoir.
Figure (18)
Figure (19)
6.4. Dynamic Equilibrium
26
DAM EQUATION OF MOTION (Zeidan 2013)
RESERVOIR EQUATION OF MOTION
6.5. Dynamic Analysis

Dynamic analysis refers to analysis of loads whose duration is short with the first period of
vibration of the structure. Such loads include seismic, blast, and impact. Dynamic methods are
appropriate to seismic loading. Because of the oscillatory nature of earthquakes, and the
subsequent structural responses, conventional moment equilibrium and sliding stability
criteria are not valid when dynamic and pseudo dynamic methods are used. The purpose of
these investigations is not to determine dam stability in a conventional sense, but rather to
determine what damage will be caused during the earthquake, and then to determine if the dam
can continue to resist the applied static loads in a damaged condition with possible loading
changes due to increased uplift or silt liquefaction. It is usually preferable to use simple dynamic
analysis methods such as the pseudo dynamic methodor the response spectrum method
(described below), rather than the more rigorous sophisticated methods. The procedure for
performing a dynamic analysis includes the following (3):
1. Review the geology, seismology, and contemporary tectonic setting.
2. Determine the earthquake sources.
3. Select the candidate maximum credible and operating basis earthquake magnitudes and
locations.
4. Select the attenuation relationships for the candidate earthquakes.
5. Select the controlling maximum credible and operating basis earthquakes from the candidate
earthquakes based on the most severe ground motions at the site.
6. Select the design response spectra for the controlling earthquakes.
7. Select the appropriate acceleration-time records that are compatible with the design response
spectra if acceleration-time history analyses are needed.
8. Select the dynamic material properties for the concrete and foundation.
27
9. Select the dynamic methods of analysis to be used.

10. Perform the dynamic analysis.
11. Evaluate the stresses from the dynamic
analysis.
6.5.1. Pseudo Dynamic Method

This procedure was developed by
Professor Anil Chopra as a hand calculated
alternative to the more general analytical
procedures which require computer programs. It
is a simplified response spectrum analysis which
determines the structural response, in the
fundamental mode of vibration, to only the
horizontal component of ground motion. This
method can be used to evaluate the compressive
and tensile stresses at locations above the base
of the dam. Using this information, degree of
damage can be estimated and factored into a post earthquake stability analysis.
Figure (20)
6.5.2. Modal Dynamic Method
Dynamic response analysis is typically performed using finite element modal analysis. The
Figure (20)
major modes of vibration are calculated, and the response of the structure to the earthquake is
expressed as a combination of individual modal responses. There are 2 acceptable techniques for
modal analysis, Response Spectrum Analysis and Time History Analysis, Figure (20).
4.3.2.1 Response Spectrum Method

In the response spectrum method, the modes of vibration determined from finite element
modeling are amplitude weighted by a response spectrum curve which relates the maximum
acceleration induced in a single degree of freedom mechanical oscillator to the oscillator's natural
period. A typical response spectrum curve is shown in figure 12. Because the timing of the peaks of
individual modal responses is not taken into account, and because peaks of all modes will not occur
simultaneously, modal responses are not combined algebraically. Modal responses are combined
using the SRSS (square root of sum of squares) or the CQC (complete quadratic combination)
methods, Figure (21).
4.3.2.2 Time History Method

The time history method is a more
rigorous solution technique. The response
of each mode of vibration to a specific
acceleration record is calculated at each
point in time using the Duhamel integral.
All modal responses are then added Figure (21)
together algebraically for each time step
throughout the earthquake event. While
this method is more precise than the
response spectrum method for a given
acceleration record, its results are
contingent upon the particulars of the
28
acceleration record used. For this

reason, time history analysis should
consider several accelerograms.
4.4.3. Direct Solution Method

The modal superposition
methods described above require the
assumption of material linearity. Direct
solution techniques solve the differential
equations of motion in small time steps
subject to material stress strain
relationships which can be arbitrary, and
therefore the development of damage
can be accounted for. Their results are
also highly affected by the particular
accelerogram used.
4.4.4 Block Rocking Analysis
When dynamic analysis techniques such as those discussed above indicate that concrete
cracking will occur, a block rocking analysis can be done. This type of analysis is useful to
determine the stability of gravity structures or portions thereof, when it is determined that cracking
will progress to the extent that free blocks will be formed. The dynamic behavior of free blocks can
be determined by summing moments about the pivot point of rocking.
4.4.6. Reservoir Added Mass
During seismic excitation the motion of the dam causes a portion of the water in the
reservoir to move also. Acceleration of this added mass of water produces pressures on the dam that
must be taken into account in dynamic analysis. Westergaard derived a pressure distribution
Figure (22)
assuming that the dam would move upstream and downstream as a rigid body, in other words, the
base and crest accelerations of the dam are assumed to be identical. This pressure distribution is
accurate to the extent that the rigid body motion assumption is valid. The dam's structural response
to the earthquake will cause additional pressure. Figure 22 shows the difference in pressure
distributions resulting from rigid body motion and modal vibration. Westergaard's theory (9) is
based on expressing the motion of the dam face in terms of a fourier series. If the acceleration of the
upstream face of the dam can be expressed as:
where α is the ground acceleration, then the resulting pressure is given by :

Figure 13
While, Westergaard assumed a rigid body acceleration, the above equations can be generalized to
accommodate any mode shape. As with the application of finite element techniques for static
analysis, the reviewer must not lose sight of the purpose of the analysis, ie to determine whether or
not a given failure mode is possible. Finite element techniques assume linear stress strain
characteristics in the materials, and almost always ignore the effect of cracking in the dam. These
assumptions can constitute rather gross errors. For this reason when reviewing the finite element
results, the stress output should be viewed qualitatively rather than quantitatively. Finite element
29
dynamic output can show where the structure is most highly stressed, but the stress values should
not be considered absolute.
4.5 Cracked Base Analysis
The dam/foundation interface shall be assumed to crack whenever tensile stress normal to
the interface is indicated. This assumption is independent of the analysis procedure used. The
practical implementation of this requirement is illustrated in the gravity analysis shown below. All
forces, including uplift are applied to the structure. Moments are taken about 0,0 which does not
necessarily have to be at the toe of the dam. The line of action of the resultant is then determined as
shown in Figure (23). The intersection of the resultant line of action and the sloping failure plane is
the point of action of the resultant on the structure. A crack is assumed to develop between the
base and foundation if the stress normal to the base is tensile. Since the gravity analysis technique
assumes a linear effective stress distribution along the dam base, the length of this crack is uniquely
determined by the location of the resultant and the assumption of a linear effective stress
distribution.
Dynamic loading is equally capable of causing base cracking, however, cracked base
analyses are not typically performed for dynamic loadings because of the computational difficulty
involved. The conventional gravity analysis procedure is not appropriate for dynamic loading
because it ignores the dynamic response of the structural system. Standard dynamic Figure
finite(23)
element
techniques are not appropriate because they are based on an assumption of material linearity and
structural continuity. What is typically assumed is that during the earthquake, extensive base
cracking does occur. Stability under post earthquake conditions, which include whatever damage
results from the earthquake, must be verified.
7. DESIGN & SAFETY CRITERIA

Specific stability criteria for a particular loading combination are dependent upon the degree
of understanding of the foundation structure interaction and site geology, and to some extent, on the
method of analysis. Assumptions used in the analysis should be based upon construction records
and the performance of the structures under historical loading conditions. In the absence of available
design data and records, site investigations may be required to verify assumptions. Safety factors are
intended to reflect the degree of uncertainty associated with the analysis. Uncertainty resides in the
knowledge of the loading conditions and the material parameters that define the dam and the
foundation. Uncertainty can also be introduced by simplifying assumptions made in analyses. When
sources of uncertainty are removed, safety factors can be lowered.
The basic requirement for stability of a gravity dam subjected to static loads is that force and
moment equilibrium be maintained without exceeding the limits of concrete, foundation or concrete
/foundation interface strength. This requires that the allowable unit stresses established for the
concrete and foundation materials not be
exceeded. The allowable stresses should be
determined by dividing the ultimate strengths of
the materials by the appropriate safety factors in
Table 2. In most cases, the stresses in the body
of a gravity dam are quite low, however if
situations arise in which stress is a concern, the
following guidance in Table 1 is applicable.
Table1:
30
Load Shear Stress on Pre-cracked Principal Axis Tension Within

Condition Failure Plane1 Intact Concrete2,4
Worst Static 1.4 σn 1.7(F'c)2/3
Max. Dynamic N.A.3 N.A.3
The tensile strength of the rock-concrete interface should be assumed to be 0. Rock foundations may
consist of adversely-oriented joints or fractures such that even if the interface could resist tension,
the rock formation immediately below may not be able to develop any tensile capacity. Therefore,
since stability would not be enhanced by an interface with tensile strength when a joint, seam or
fracture in the rock only a few inches or feet below the interface has zero tensile strength, no tension
will be allowed at the interface.
7.1. Sliding Stability Safety Factors

Recommended factors of safety are listed in Table 2
Table 2
Recommended Minimum Factors of Safety 1/
 Dams having a high or significant hazard potential.
 Loading Condition 2/ Factor of Safety 3/
 Usual 3.0
 Unusual 2.0
 Post Earthquake 4/ 1.3
 Dams having a low hazard potential.
 Loading Condition Factor of Safety
 Usual 2.0
 Unusual 1.25
 Post Earthquake Greater than 1.0
Notes:
1. Safety factors apply to the calculation of stress and the Shear Friction Factor of Safety within
the structure, at the rock/concrete interface and in the foundation.
2. Loading conditions as defined in paragraph 3-3.0.
3. Safety factors should not be calculated for overturning, i.e., Mr / M0.
4. For clarification of this load condition.
One of the main sources of uncertainty in the analysis of gravity dam stability is the amount
of cohesive bond present at the dam foundation interface. The FERC recognizes that cohesive bond
is present, but it is very difficult to quantify through borings and testing. It has been the experience
of the FERC that borings often fail to recover intact interface samples for testing. In addition,
strengths of intact samples that are recovered exhibit extreme variability. For this reason, Table 2A
below offers alternative recommended safety factors that can be used if cohesion is not relied upon
for stability.
Table 2A
 Alternate Recommended Minimum Factors of Safety
 for Use in Conjunction with a No Cohesion Assumption
 Loading Condition Factor of Safety
 Worst Static Case 5/ 1.5
 Flood if Flood is PMF 6/ 1.3
 Post Earthquake 1.3
Notes:
31
1. The worst static case is defined as the static load case with the lowest factor of safety. It
shall be up to the analyst to determine the worst static case and to demonstrate that it truly is
the worst static case.
2. Because the PMF is by definition the flood that will not be exceeded, a lower factor of safety
may be tolerated. Therefore if the worst static case is the PMF, a factor of safety of 1.3 is
acceptable. If the IDF is not the PMF, then the safety factor for the worst static case shall
control.
6.2. Cracked Base Criteria

For existing structures, theoretical base cracking will be allowed for all loading conditions,
provided that the crack stabilizes, the resultant of all forces remains within the base of the dam, and
adequate sliding safety factors are obtained. Cohesion may only be assumed on the uncracked
portion of the base. Limitations may be necessary on the percentage of base cracking allowed if
foundation stresses become high with respect to the strength of the concrete or foundation material.
When remediation is required, the remediation should be designed to attempt to eliminate
theoretical base cracking for static load cases.
6.3 Safety Factor Evaluation

The safety factors determined in accordance with the previous sections shall be evaluated on
a case-by-case basis in order to assess the overall safety of a particular project. Engineering
judgment must be used to evaluate any calculated safety factor which does not conform to the
recommendations of tables 1, 2 . In applying engineering judgment, consideration must be given to
both the adequacy of the data presented in support of the analyses and the loading case for which the
safety factor does not meet the criteria. It is preferable to conservatively define strength parameters
and loading conditions than to utilize higher safety factors to accommodate uncertainties in the
analysis. Therefore, if the analyst can demonstrate that there is sufficient conservatism in the
strength parameters and analysis assumptions, lower factors of safety may be considered adequate
on a case by case basis. Any decision to accept safety factors lower than those shown in Table 2A of
this chapter will be based on: (1) the degree of uncertainty in the data and analyses provided and (2)
the nature of the loading condition, i.e. its probability of exccedance.
In accepting any lower safety factor as outlined herein, the stability analyses must be
supported by a program that includes, but is not limited to, adequate field level investigations to
define material (dam and foundation) strength parameters, installation and verification of necessary
instrumentation to evaluate uplift assumptions and loading conditions, a detailed survey of the
condition of the structure, and proper analysis procedures. This program should be submitted for
approval by the Director, Division of Dam Safety and Inspections. Flexibility on safety factors
beyond that discussed above will be infrequent and on special case-specific consideration.
6.4 Foundation Stability

The foundation or portions of it must be analyzed for stability whenever the structural
configuration of the rock is such that direct shear failure is possible, or whenever sliding failure is
possible along faults, shears and/or joints. Associated with stability are problems of local over
stressing in the dam due to foundation deficiencies. The presence of such weak zones can cause
problems under either of two conditions: (1) when differential displacement of rock blocks occurs
on either side of weak zones, and (2) when the width of a weak zone represents an excessive span
for the dam to bridge over. Sliding failure may result when the rock foundation contains
discontinuities and/or horizontal seams close to the surface. Such discontinuities are particularly
dangerous when they contain clay, bentonite, or other similar substances, and when they are
32
adversely oriented. Appropriate foundation investigation and exploration must be done to identify
potential adverse features.
Gravity dams constructed on soil foundations are usually relatively small structures which
exert low bearing pressures upon the foundation. Large structures on soil foundations are usually
supported by bearing or friction piles. When the foundation consists of pervious sands and gravels,
such as alluvial deposits, two possible problems exist; one pertains to the amount of underseepage,
and the other is concerned with the forces exerted by the seepage. Loss of water through
underseepage may be of economic concern for a storage or hydroelectric dam but may not adversely
affect the safety of the dam. However, adequate measures must be taken to ensure the safety of the
dam against failure due to piping, regardless of the economic value of the seepage. The forces
exerted by the water as it flows through the foundation can cause an effective reduction in the
weight of the soil at the toe of a dam and result in a lifting of the soil. If uncontrolled, these seepage
forces can cause a progressive erosion of the foundation, often referred to as "piping" and allow a
sudden collapse of the structure.
The design of the erosion, seepage and uplift control measures requires extensive knowledge
of type, stratification, permeability, homogeneity, and other properties of the foundation materials.
One way to limit this type of material transport is to insure that the weighted creep ratio is greater
than the minimum values shown in Table 3.
Table 3
Minimum Weighted Creep Ratios, Cw for Various Soils
 Very fine sand or silt 8.5
 Fine sand 7.0
 Medium sand 6.0
 Coarse sand 5.0
 Fine gravel 4.0
 Medium gravel 3.5
 Coarse gravel including cobbles 3.0
 Boulders with some cobbles 2.5 and gravel.
Some of the control measures which may be required may include some, all or various
combinations of the following devices:
 Upstream apron, usually with cut offs at the upstream end.
 Downstream apron, with scour cut offs at the downstream end, and with or without filters
and drains under the apron.
 Cutoffs at the upstream or downstream end or at both ends of the overflow section, with or
without filters or drains under the section.
8. SIESMIC ANALYSIS OF CONCRETE GRAVITY DAMS

8.1. Earthquake Analysis of Concrete Gravity Dams
In order to design earthquake resistant dams and evaluate the safety of existing dams that
will be exposed to future earthquakes, it is essential to have accurate and reliable analysis
procedures to predict the stresses and deformations in dams subjected to earthquake ground motion.
For a dam-water-foundation system, the earthquake response is significantly influenced by the
interaction of the dam with the impounded water and with the underlying foundation region, thus
increasing the requirements for the analysis procedure to be used, and complicating what would
otherwise have been considered a routine finite element analysis of a concrete cross-section.
33
A response history analysis (RHA) procedure, based on the substructure method, was
presented in 1981 to determine the earthquake response of concrete gravity dams including the
hydrodynamic effects of the impounded water and the effects of interaction between the dam and a
flexible foundation. In 1984, this RHA procedure was extended to also recognize absorption of
hydrodynamic pressure waves into the alluvium and sediments invariably deposited at the bottom of
reservoirs. Through a comprehensive investigation it was shown that the effects of dam-water-
foundation interaction and reservoir bottom absorption has a profound influence on the response of
concrete gravity dams to horizontal and vertical ground motion. The above-mentioned analysis
procedure was implemented in the computer program EAGD-84 to numerically evaluate the
response of a two-dimensional dam-water-foundation system to earthquake ground motion (37).
The safety of dams during earthquakes is extremely important because failure of such a
structure can have catastrophic consequences on life and property. It is therefore essential to have
reliable analysis procedures to design earthquake resistant dams and evaluate the safety of existing
dams. Traditional "static" design procedures have been widely used to design concrete dams – and
are in some cases still being used – even though it has been shown repeatedly that they are based on
unrealistic assumptions, and that dams designed according to these procedures have experienced
widespread damage during earthquakes.
In 1978, a two-stage procedure was proposed for the elastic analysis phase of seismic design
and safety evaluation of concrete gravity dams: (1) response spectrum analysis (RSA) in which the
peak value, i.e., the maximum absolute value, of response is estimated directly from the earthquake
design spectrum; and (2) response history analysis (RHA) of a finite element idealization of the dam
monolith. The RSA procedure was recommended for the preliminary phase of design and safety
evaluation of dams and the RHA procedure for accurately computing the dynamic response and
checking the adequacy of the preliminary evaluation. In the mid 1980's, both procedures were
extended to consider the full effects of dam-water interaction, dam-foundation interaction and
reservoir bottom absorption, known to have profound influence on the response of a dam to
earthquake ground motion. Both the RHA and RSA procedures have been implemented in computer
software, which have been utilized extensively for research purposes and in actual projects (38).
8.2. Response Spectrum Analysis Procedure

The response spectrum analysis (RSA) procedure developed to estimate the earthquake
induced stresses in concrete gravity dams considers only the more significant aspects of the
response. Although the dynamics of the system including dam-water-foundation interaction is
considered in estimating the response due to the fundamental vibration mode, the less significant
part of the response due to higher modes is estimated by the static correction method. Only the
horizontal component of ground motion is considered because the response due to the vertical
component is known to be much smaller. Dam-water-foundation interaction introduces frequency-
dependent, complex-valued hydrodynamic and foundation terms in the governing equations. Based
on a clever series of approximations, frequency-independent values of these terms were defined and
an equivalent SDF system developed to estimate the fundamental mode response of dams, leading
to the RSA procedure summarized in the subsequent sections. This development was presented and
approximations evaluated and justified in a series of publications.
The two-dimensional system considered consists of a concrete gravity dam monolith
supported on a horizontal surface of underlying flexible foundation rock, idealized as a viscoelastic
half-plane, and impounding a reservoir of water, possibly with alluvium and sediments at the
bottom (37).
34
8.3. Hydrodynamic Pressure Analysis

During earthquake occurrence, the dam and reservoir body respond differently, as a result of
hydrodynamic forces impinging on the fluid body and solid structure. As a result of this, interaction
will occur between the fluid–solid structure interfaces as particles move relatively to the mesh
points whereas, the meshes moves with the material particles. Much research work has been carried
out for the dynamic response of the fluid solid structure systems. Several methods of analysis for the
fluid structure systems use finite element idealization in the nonlinear dynamic response of the
system. Earthquake-induced hydrodynamic pressures on upstream face of a dam are important
factors in design consideration. Assuming that the fluid is incompressible, Westergaard [9] was the
first who derived an analytical solution for the hydrodynamic pressure acting on a rigid dam with a
vertical upstream face as a result of horizontal harmonic ground motion. In this method for the
analysis of concrete dams, fluid is treated as an added mass to the body of the dam. In the last fifty
years, many researchers have extended the Westergaard’s classical work to include more physical
parameters such as the compressibility of the fluid in the reservoir, the flexibility of the dam, and
reservoir bottom absorption. However, analytical solutions are rare and available only for a reservoir
with a simple geometry and boundary conditions. Numerical models proved to be powerful
techniques that can be used for dam-reservoir interaction problems with complex two and three
dimensional geometries as well as complicated boundary and initial conditions.
During earthquake occurrence, the dam and reservoir body respond differently, as a result of
hydrodynamic forces impinging on the fluid body and solid structure. As a result of this, interaction
will occur between the fluid–solid structure interfaces as particles move relatively to the mesh
points whereas, the meshes moves with the material particles. Several research work has been
carried out for the dynamic response of the fluid-solid structure systems. Many researchers have
paid attention for the dynamic response of the foundation-solid structure system. Dam-reservoir-
foundation interaction of concrete gravity dams subjected to ground motions have been addressed
(8).
9. RISK ANALYSIS FOR CONCRETE GRAVITY

STRUCTURES
For concrete gravity dams founded on alluvial soils, the leading cause of failure is piping or
―blowout‖ of the soil material from beneath the dam. Therefore, the reader is referred to the section
on Internal Erosion and Piping Risks for Embankments for evaluating this potential failure mode,
considering ―backward erosion piping‖ of the foundation soils. The heel of the dam is a location of
sharp geometry change and as such is a point of singularity and stress concentration. Thus, the dam-
foundation contact is typically the focus of most of the stability analysis. However, this typically is
not the weak link in the dam-foundation system, unless the dam is founded on the foot wall of
smooth discontinuity surfaces such as faults or bedding planes. The rough surface that results from
blasting the dam keyway excavation typically provides a significant roughness or ―dilation‖
component to the shear strength on this surface, which should be taken into account to the extent
possible based on construction photographs and other information. If the surface clean-up is good,
significant cohesion and tensile strength can result (as with lift joints).
When surface cleanup of lift joints is not good, weaker horizontal planes may occur within
the dam body. For gravity dams constructed in blocks, the weaker planes may not ―line up‖ across
contraction joints, and if the joints are constructed with keys, considerable stability can result from
load transfer to adjacent monoliths. This should be considered when evaluating the risks A line of
functioning drainage holes in the foundation or dam body adds significantly to the sliding stability
of concrete gravity dams by reducing water pressures (typically referred to as ―uplift‖) along
potential sliding surfaces. A decrease in water pressures increases the effective normal stress and
35
frictional resistance. Research shows that drains remain effective even if a crack or open surface
extends downstream of the drainage curtain as noted in nonlinear analysis guidelines, based on the
Electric Power Research Institute (EPRI) research results. However, drainage systems can become
plugged over time if they are not maintained, and the drainage curtain can be offset under significant
seismic displacements, thus becoming less effective.
Shear keys constructed within the contraction joints separating concrete monoliths are
beneficial in that they can facilitate load transfer between monoliths. This could be important if one
monolith or series of monoliths contains an unbounded lift joint or weak foundation conditions,
whereby load in excess of the weak monolith(s) capacity could be transferred to adjacent stronger
monoliths. Not all gravity dams contain shear keys within the contraction joints, but many do.
When a potential sliding plane is formed by a partially bonded and partially unbounded surface, care
must be taken in assigning the shear strength to each portion. That is because the peak shear
strengths may not be mobilized at compatible displacements. It may take much less shear
displacement to mobilize the shear strength of a bonded joint than an unbounded joint, in which
case it may not be appropriate to simply add the peak strengths determined from testing. Test results
could be examined and new strength curves developed at compatible displacements (39).
9.1. Tensile Strength of Concrete

The tensile strength of concrete has been somewhat controversial over the years. Jerry
Raphael published an often cited paper on the subject (Raphael, 1984). His basic conclusions were
that:
• Direct tension tests are unreliable, and can be in error by as much as 50 percent (attributed to
moisture gradients during drying which caused surface micro-cracking and an effective reduction on
cross-sectional area in these types of tests – Cannon (1995) noted that the drilling process can also
induce strains capable of causing micro-cracking of the core surface).
• Splitting tension tests are the most reliable means of determining tensile strength (potential zone of
micro-cracking is loaded in compression).
• Tensile strength determined by static testing (from splitting tension tests) should be increased by
50 percent when used with seismic loadings, based on rapid loading tests where the samples are
taken to failure in a fraction of a second representing one load cycle during an earthquake.
Raphael supported the fact that splitting tensile tests best represent tensile strength by
converting strengths from modulus of rupture (flexural) tests results to tensile strength based on an
evaluation of the stress and strain distribution in the samples, which shows the tensile strength
should be about ¾ the modulus of rupture value. The results, suggest that both splitting tension tests
and modulus of rupture tests produce a consistent pattern. Raphael indicated that an apparent tensile
strength can be used when performing linear elastic analyses to account for the non-linear strain that
occurs prior to failure. The apparent strength is estimated as the failure tensile strain multiplied by
the Modulus of Elasticity used in the analysis. It can be determined directly from modulus of rupture
(flexural) tests, or can be estimated from equations he provides (which result in approximately a
factor of 1.35 applied to the static strengths). However, this typically has not been used in practice
due to discomfort with using the high strengths it produces, particularly for static loading. However,
for dynamic loading, it can be considered, especially if it is taken that only one spike in the stress
time history is sufficient to crack the concrete (considering that it represents a rapid load cycle
similar to that experienced under dynamic testing).
Slightly over a decade later, Bob Cannon published additional information in a Corps of
Engineers Engineering Pamphlet (Cannon, 1995). He examined direct tension and splitting tension
test results for a variety of conventional concretes, and concluded that:
• Splitting tensile strengths can be used as a starting point. If these strengths are not available, then:
(1) for compressive strengths less than 3,000 lb/in2 the tensile strength is expected to vary between
36
10 and 15 percent of the compressive strength, or (2) for compressive strengths greater than 3,000
lb/in2, the equations 1.7(fc’)2/3 (Raphael’s equation) or 7(fc’)1/2, which are based on splitting test
relationships, can be used to estimate tensile strength.
• The tensile strength values should be reduced by 10 percent if the maximum size aggregate is
larger than 1½ inches (based on 6-inch x 12-inch cylinders).
• The strengths should be reduced by an additional 20 percent to adjust for direct tensile strength
(particularly for examining vertical or ―cantilever‖ stresses).
Cannon also gave recommendations for roller-compacted concrete (RCC) indicating that:
• Parent material tensile strength should be no higher than about 75 percent of the splitting tensile
strength value, reduced by 10 percent if based on wet-screening of aggregates larger than 1½ inches.
• Joint tensile strength is similar to conventional concrete when properly cleaned, cured, and
covered with a suitable mortar or bedding mix.
Cannon supported Raphael’s conclusions that:
• The dynamic tensile strength of concrete is about 1.5 times the static tensile strength.
• For linear elastic finite element analysis, the apparent tensile strength of concrete is about 1.35
times the tested strength.
The intent of reviewing these important pieces of work is not to dictate the tensile stress
parameters that should be used in a risk analysis. These need to be determined on a case-by-case
basis using available information. However, the work by Cannon is important and often overlooked
in estimating tensile strength. It is always preferred to have tested material properties from extracted
core from the dam. The direct and splitting tensile strengths can significantly vary from dam to dam
at shown in the Non-Linear Practices Manual (Mills-Bria et al, 2006). Lift line strength is not only a
function of the concrete strength but is greatly influenced by construction methods. Experience
suggests that tensile strength across lift joints for modern concrete construction with good joint
clean-up averages about 85 percent of the parent concrete strength. Good cleanup usually involves
water curing the top of new concrete lifts, then ―green cutting‖ or water blasting (sometimes
sandblasting) the laitance from the top of a lift prior to placing the overlying concrete. Sometimes a
layer of mortar or richer concrete with smaller aggregate is placed first to bond to the underlying lift.
Lower strengths could be present for concrete where lift clean-up and material quality control was
questionable.
Seepage on the downstream face of a concrete dam is not a reliable indicator of lift joint
bond. Friant Dam has many large seeps on the downstream face along the lift lines but extracted
core indicated bonded lift surfaces. It appeared that the bottom of the concrete lifts were not
consolidated as well and were more porous than the overlying concrete, forming a seepage pathway,
but that enough paste and contact was maintained to bond to the underlying lift. In contrast, Stewart
Mountain Dam has a dry downstream face, but extracted core showed 16 of 23 lift joints unbonded.
Failure to thoroughly clean the laitance from the lift surfaces resulted in weak bond, but there were
sufficient fine particles at the interface to limit seepage along the joint. Therefore, it is important to
locate as much information as possible about the methods and specific conditions encountered
during construction, which are often keys to the strength of lift joints. If insufficient information can
be located to make a judgment on lift joint strength at the time of the risk analysis, best practice is to
perform some analysis results with poorly or unbonded lifts to judge the stability implications if this
condition exists. It only takes one poorly bonded lift to create a potentially high risk situation. It
should be noted that any empirical relationships between concrete compressive strength and tensile
strength do not apply to concrete that has been affected by alkali aggregate reaction (AAR). AAR
results in formation of a gel around the aggregate particles. Therefore, while the compressive
strength of the concrete may remain at a fairly high level, the tensile strength is often greatly
reduced. Site specific testing is typically required in this case (39).
37
9.2. Cracked Base Analysis

The ―cracked base‖ analysis has found its way into most concrete gravity dam design
criteria, based on the ―gravity‖ method of analysis, which assumes plane sections remain plane, and
thus the distribution of vertical stress is linear. It is often applied without thoroughly evaluating the
reasonableness of the results or the analysis assumptions relative to actual conditions. In a risk
context, these must be considered. Several important points in this regard include:
• There is often confusion in how to deal with total stress and effective stress in carrying out the
calculations. Design of Small Dams (1987) indicates that ―Uplift from internal water pressures and
stresses caused by the moment contribution from uplift along a horizontal plane are usually not
included in the computation of σZ.‖ This is the total stress method, which is endorsed by
Watermeyer (2006), who states that the ―reactive stress equations [which include the contribution
from uplift] are erroneous and can lead to erroneous conclusions when uplift reducing drains are
incorporated into the base of a gravity dam.‖ That is not to say that uplift is not considered in the
analysis, only that the moment contribution from internal uplift forces are not included in the stress
calculations.
• The effective stress is determined by subtracting the pore water pressure (often equated to the
―uplift pressure‖) from the total stress. If the effective stress is tensile and exceeds the tensile
strength, then it is assumed that cracking can initiate. At that point, the water force in the crack
becomes an ―external‖ force which is included in the total stress calculations, and the base length is
assumed to be shortened to only that portion downstream of the crack tip. The effective stress at the
crack tip is subsequently calculated as the difference between the total stress and effective stress at
that location. It should be noted that the crack may not progress downstream of the point at which
the effective stress is equal to the tensile strength.
• At the base of the dam, the potential for full reservoir pressure at the crack tip is controlled by the
permeability of the foundation. Concrete gravity dams are typically founded on fractured and jointed
rock. Thus, full reservoir pressure cannot develop at the tip of a crack along the foundation contact
unless the foundation rock is massive and un-fractured, or the foundation joints are much tighter
than the base crack. This is because water entering the crack will flow out through fractures at the
base of the dam, and head loss will occur due to this flow. Thus, full uplift in a crack tip at the
foundation contact may not be reasonable.
• Drains remain effective even if penetrated by a horizontal crack, although the drain efficiency may
be reduced somewhat. This is demonstrated by the research sponsored by the Electric Power
Research Institute at the University of Colorado (Amadei et al, 1991). Thus, analyses which
consider full hydrostatic reservoir pressure in a crack tip downstream of the line of drains are
typically not used for risk analyses.
• In the limiting case, if a crack is judged to propagate completely through the structure, the uplift
pressure distribution along the crack is that which is appropriate for the post-cracking conditions,
including the effects of drains in reducing the pressures, and pressures no higher than tailwater at
the downstream face. It should be noted that there is very little guidance currently available
concerning the effects of drains if the section cracks all the way through. If the aperture of the crack
is thought to remain relatively tight in comparison to the drain diameter, the drains should retain
some effectiveness. If the aperture is thought to be large in comparison to the drain diameter, then
there may be more flow than the drains can handle, and their effectiveness would be questionable.
• If a crack is shown to exist, cohesion is presumed to act only on the portion of the intact potential
sliding plane that is in compression. It is expected that intact concrete in tension will exhibit a
smaller cohesive strength component, and since this is difficult to quantify, it is typically ignored.
38
9.3. Risks under Normal Operations

Concrete gravity dams that have performed well under normal operating conditions will
likely continue to do so unless something changes. Changes could result from plugging of drains
leading to an increase in uplift pressures, possible gradual creep that reduces the shear strength on
potential sliding surfaces, or degradation of the concrete from alkali-aggregate reaction, freeze-
thaw, or sulfate attack. These may be difficult to detect. A review of instrumentation results can be
helpful. For example, if piezometers or uplift pressure gauges indicate a rise in pressures, and weirs
indicate a reduction in drain flows, the drains may be plugging leading to potentially unstable
conditions. If conditions appear to be changing, risk estimates are typically made for projected
conditions as well as current conditions.
Reliability analysis for sliding on near horizontal foundation planes and/or potentially weak
or cracked horizontal lift joints, typically using two-dimensional analysis sections, is the primary
tool used for estimating risks posed by concrete gravity dams under normal operating conditions.
This involves performing a probabilistic stability analysis using the Monte-Carlo technique as
described in the section on Reliability Analysis. It requires an assessment of the likely range in input
parameters, such as drain efficiency, cohesion and friction coefficient along the potential sliding
surface, percentage of potential sliding surface that is intact, orientation of the potential sliding
surface, and unit weight of the material(s). For potential foundation sliding planes, the influence of a
downstream passive rock wedge should be considered, where appropriate. The shear strength of
rough surfaces is nonlinear as a result of ―riding up‖ over asperities at low normal stress and
shearing through them at high normal stress. A straight line fit through such data points can result in
overestimating the shear strength, particularly at low normal stresses. Therefore, strength parameters
should be selected for the appropriate normal stress range of interest, or other means used to account
for the nonlinear shear strength envelope.
Probabilistic stability analyses are typically performed at various reservoir water surface
elevations, and combined in an event tree, such as that shown in Figure 20-1. See the sections on
Event Trees and Reservoir Level Exceedance Curves for information on calculating reservoir load
range probabilities. Note that in the limit, if small enough reservoir elevation increments are
selected, a curve, referred to as a ―fragility curve‖, results. The calculations are essentially the same
whether larger discrete ranges or a fragility curve is used, and the results are similar as long as care
is taken in selecting the discrete ranges. Therefore, either method can be used in estimating risks.
For the probabilistic stability analyses, it is important to examine the sensitivity rank coefficients
and perform parametric studies, varying the parameters that affect the results the most. These
parametric studies are used to estimate an appropriate range in conditional failure probabilities for
the node titled ―Sliding Instability‖. If there are significant three-dimensional effects, the two-
dimensional sliding model may not be appropriate, and three-dimensional analyses may be needed
to get a handle on how significant these effects might be if risks estimated from the two-
dimensional models exceed the public protection guidelines.
9.4. Risks under Flood Loading

The approach for estimating risks due to structural instability under flood loading is essentially the
same as for static loading, except that reservoir water surface elevations above the normal operating
range, assigned the appropriate flood frequency, are used in the analyses and event tree. If flood
routing information is not available, a conservative initial assumption is that inflow is equal to
outflow, and the level of the reservoir is determined by that needed to pass a given peak inflow
through the spillway and/or other release facilities (see also the section on Dam Overtopping). If the
risks using this method are in an area where risk reduction actions are justified, then flood routings
may be needed to get a better handle on the probability of attaining various reservoir elevations.
39
As the reservoir rises during flood loading, there may be a level at which the heel of the dam
goes into tension (based on effective stress), in which case the potential for cracking along a lift
joint at that elevation may increase. At some point, the estimated tensile strength of the concrete
may be exceeded. Typically, a separation in the event tree reservoir load ranges occurs at these
reservoir elevations. Stability analyses should be performed at these reservoir water elevations to
judge the impact on the dam. Make sure the tailwater and uplift conditions correspond to the given
reservoir elevation. In the case of an overflow section, care must be taken when assuming nappe
forces (forces due to water flowing above the spillway) and tailwater forces act on the dam. Stilling
basins can ―sweep out‖ at high flows, and nappe pressures can become subatmospheric, reducing
the stabilizing forces. Forces generated by water flowing through a flip bucket can also affect the
results. A hydraulic evaluation is typically performed to determine whether it is appropriate to
include these forces. A reliability model with the proper formulation for a cracked base analysis (see
Watermeyer, 2006) is important in examining conditions where tension exceeding the tensile
strength develops.
Risk evaluation associated with overtopping erosion of the abutments or foundation is
discussed in the sections on Flood Overtopping and Erosion of Rock and Soil. However, another
potentially significant issue involves cases where a concrete gravity dam serves as a spillway
section. If erosion occurs at the downstream toe of the structure during spillway releases, weak
bedding planes or foundation discontinuities in the underlying foundation rock might be exposed,
daylighting into the erosion hole. This could remove passive resistance from the downstream rock
mass, and result in a much more unstable condition. See the section on Erosion of Rock and Soil for
guidance on how to estimate the potential for erosion. Figure 20-2 shows how this might impact the
event tree. The potential for failure of stilling basins is discussed in the section on Overtopping of
Walls and Stilling Basin Failure.
9.5. Risks under Earthquake Loading

Under earthquake loading, concrete gravity dams will respond according to the level and
frequency of the shaking, and the reservoir level at the time of shaking. Therefore, sufficient
analyses need to be performed to evaluate conditional failure probabilities at various levels of
shaking and reservoir elevation. An example event tree to examine the potential for sliding failure
through a weak lift line at a sharp change in slope on the downstream face is shown in Figure 20-3.
For each reservoir and seismic load range that is established for the estimating process, the
likelihood of cracking through the dam body at this location must be estimated. The best approach
for this is to perform a nonlinear dynamic finite element studies, modeling the potential weak plane
with a contact surface that can be assigned a tensile strength value. As the tensile strength is
exceeded near the faces during seismic response, the nodes will separate. If the shaking is severe
enough, complete separation of the contact surface may propagate through the structure. Figure 20-4
shows a horizontal contact surface through a three-dimensional model of a concrete gravity dam.
The darker color represents portions that remained un-cracked following the earthquake shaking.
This indicates that at least one monolith cracked completely through for the set of assumptions used
in this analysis. Similar studies can be performed using a two-dimensional section. By varying the
tensile strength within reasonable parameters and monitoring the percentage of the joint that
separates, a range in the likelihood of complete separation can be made. It should be noted that this
is a total stress analysis, and pore pressures are not considered. Pore pressure behavior in concrete
under dynamic loading is a subject of much uncertainty. Therefore, it is typically assumed that the
total stress analysis provides a reasonable approximation of the potential for cracking through the
section.
If the dam only cracks partially through, the probability of post-earthquake instability in the
estimated cracked state is determined using static reliability analysis, as previously described. The
40
estimated crack length from the nonlinear analysis of the seismic shaking is used as the starting
point for a cracked base analysis. It is very difficult to estimate the amount and depth of cracking
from a linear analysis. Linear analyses only help determine if and where cracks might initiate (high
stress areas) but cannot model crack development or the sudden release of kinetic energy when
cracks form.
If the section cracks all the way through, the likelihood of shearing the drains is next
estimated. Information typically used to make this assessment includes calculated displacements
from the finite element study assuming frictional resistance only on the potential sliding surface, as
shown in Figure 20-5. In this case, very small values of damping, only enough to keep the model
stable as the loading is applied, need to be used. If the model is over-dampened, the displacements
will be under-estimated. Although this type of analysis assumes the section is cracked at the
beginning of the earthquake and thus are somewhat conservative, they can be used to estimate the
likelihood of drains, where present, being sheared. The post-earthquake instability could be
considerably different whether the drains are still functioning after the earthquake shaking or not. It
is possible that the drains could be sheared off, or opening of pathways in the foundation could lead
to increased flow that overwhelms the drainage system. Therefore, two estimates are made, using
reliability analysis, to account for these two conditions (drains functional or not), as indicated by the
nodes on the event tree in Figure 20-3.
Seismic risk analysis of concrete gravity dams typically relies heavily on finite element
analyses to evaluate the dynamic response, and the ―gravity method‖ analyses to evaluate post-
earthquake stability. The finite element analyses described above are not routinely performed.
Although more uncertain, if analyses that include a contact surface are not available, it may be
necessary to make judgments on cracking from traditional linear elastic finite element analysis
results, by examining the magnitude and duration of the vertical tensile stresses at the upstream and
downstream faces. Judgments must be made concerning how load is redistributed if cracking begins
at the face, and how far toward the center of the dam it will progress, which is not an easy task. It is
also important to examine the three-dimensional effects and, for example, whether excess driving
load can be transferred to adjacent monoliths through shear keys. This is particularly true if all
analyses are based on two-dimensional sections.
9.6. Accounting for Uncertainty

Uncertainty is accounted for by estimating a range or distribution of values for each node on
the event tree. A Monte-Carlo analysis is then run for the event tree to display the ―cloud‖ of
uncertainty, as described in the section on Combining and Portraying Risks. It is important to
perform parametric or sensitivity analyses to examine how the results might change with different
input parameters, especially for reliability analyses as described in the section on Probabilistic
Stability Analysis. Different assumptions on the distribution and magnitude of water forces
following an earthquake are typically made, since there is typically a great deal of uncertainty
surrounding these values, and they can have a controlling effect on the results of the analyses. The
uncertainty associated with how well the models are thought to actually predict the complex
behavior should also be factored into the estimates, perhaps in a parametric sense (i.e. vary the
corrections to account for model uncertainty and examine the results on the risk estimates).
9.7. Probabilistic Seismic Risk Assessment

Seismic resistance of concrete gravity dams is in general greater than that of other common
types of dams or structural systems. Their massive and compact structure provides large structural
stability, making an instantaneous collapse due to earthquake loading practically impossible. For
example, just two of the 18 spillways of the Shih-Kang concrete gravity dam collapsed during the
41
1999 Chi-Chi earthquake (M=7.3) due to the extremely large fault movements. Collapse of concrete
gravity dams is therefore not instantaneous, which may help to mitigate a disaster of high
proportions in the case of major earthquake. However, for earth gravity dams an instantaneous
collapse was observed. The biggest disaster caused by dam failure probably happened in China in
1975. The collapse of earth dam Banqiao caused deaths of 26000 people and indirectly additional
145000 due to famine and epidemics. More than 6 million buildings were destroyed. The cause of
collapse was erosion of the dam as a consequence of extreme weather conditions. The water spilled
over the crest of the dam causing erosion and consequently collapse. Such failure mode is
impossible in the case of concrete gravity dams.
Significant damage of concrete gravity dams due to earthquakes is rarely observed. Strong
earthquake that may cause considerable damage to concrete gravity dams usually cause many
fatalities which are not a consequence of the collapse of a dam, but are the ordinary buildings which
collapse for such strong earthquakes. This happened in the case of the 2008 Sichuan earthquake.
The earthquake had a magnitude of 7.9 MMS and caused the death of more than 80000 people. A
number of concrete dams were built in the area and some of them were seriously damaged. Despite
the formation of large cracks, none of the dams collapsed, neither a significant spill of water was
observed. Therefore the victims and the damage on surrounding buildings were not caused due to
failure of dams. It should be mentioned that an earthquake of such magnitude is not likely to occur
in Slovenia. For example, during the 2008 Sichuan earthquake approximately a thousand times
larger amount of energy was released comparing to 1998 earthquake in Posočje.
Therefore a question arises whether the seismic risk for losses due to damage or
collapse of concrete gravity dams in Slovenia is tolerable or not? Several approaches are available
to assess the seismic safety of concrete gravity dams (e.g. [3]), but, in general, the assessment of
seismic safety is associated with uncertainties which are of aleatoric (e.g. ground motions) or
epistemic nature (lack of knowledge). Therefore the problem should be treated in a probabilistic
manner. A brief overview of simplified seismic risk assessment methodology and the description
between the deterministic intensity-based assessment (or design) and probabilistic seismic risk
assessment of structures is firstly presented in the paper. The methodology, which could be used in
practice to assess seismic risk of existing or newly designed concrete gravity dams, is then
demonstrated by means of an example of concrete gravity dam Učja (40).
9.8. Fracture Analysis of Concrete Gravity Dams

Dynamic crack propagation analysis was used to assess the seismic safety of concrete gravity
dams in large-scale earthquakes. Since damping characteristics have been shown to have a major
effect on the occurrence and propagation of cracks in dam body, we first discuss damping
characteristics by comparing existing shaking table test results with the results of dynamic crack
propagation analysis. We then perform dynamic crack propagation analysis on a full-scale dam
model to make clear crack occurrence and propagation during a large-scale earthquake. By
performing dynamic crack propagation analysis while varying the acceleration level of the input
seismic motion, we define the residual factor of ligament length as a new measure related to the
crack propagation length, and we use it to perform a quantitative evaluation of seismic safety
against crack penetration failure. Since a concrete gravity dam is a plain concrete structure, the
ultimate state under large-scale earthquakes is expected to be penetration failure resulting mainly
from the occurrence and propagation of cracks in the dam body. Various analytical studies have
already looked into the propagation behavior of cracks in dam body. Factors that influence the
occurrence and propagation of cracks in dam body include the input seismic motion, the amount of
water in the reservoir (i.e., the water level), and tension softening characteristics of concrete and
damping characteristics. In earlier studies, Zhang and Ohmachi studied the effects of the reservoir
42
water and the penetration of water into cracks (Zhang and Ohmachi (1998); Zhang and Ohmachi
(2000)), and Sasaki et al. studied how the tension softening characteristics of concrete affect
properties such as the failure energy and tensile strength (Sasaki and Kanenawa (2003); Sasaki et
al, (2003)).
Damping characteristics have been shown to have a particularly large effect on the
occurrence and propagation of cracks in the dam bodies of concrete gravity dams. At the Ministry of
Economy, Trade and Industry (METI), in order to make clear the crack propagation behavior of
concrete gravity dams during large-scale earthquakes, two-dimensional FEM nonlinear dynamic
analysis has been performed using a smeared crack model that takes the tension softening
characteristics of the concrete into account. However, this technique has been shown to be incapable
of getting the crack localization phenomenon inherent in plain concrete structures through the
distribution of cracking, and there are problems involved in a damping evaluation method (METI
and JEPOC (2001)). El-Aidi and Hall (1989) [1,2] showed that damping characteristics have a large
effect on crack propagation behavior, and that the mass proportional parts in Rayleigh damping
constitute a resisting force at the surfaces where cracks occur so that the stress release becomes
insufficient. They applied stiffness-proportional damping to a nonlinear dynamic analysis.
Bhattacharjee et al. (1993) and Horii and Chen (2003) also reported similar findings, and in general
it is thought that crack localization phenomena can be indicated by using the initial stiffness to set
the stiffness-proportional damping but employing the time-domain stiffness proportional damping
that degrades from one moment to the next as cracks propagate. Also, Lee et al. (1998) evaluated
the crack propagation behavior in dams by introducing a damage variable to vary the damping
characteristics according to the amount of damage (the stiffness softening associated with crack
propagation).
Since these studies have shown that the damping characteristics have a significant effect on
the occurrence and propagation of cracks in the dam bodies of concrete gravity dams, in the present
study we first investigated a suitable method for evaluating damping characteristics by comparing
the results of dynamic crack propagation analysis and shaking table tests (Tinawa et al. (2000)) on
plain concrete structures made to simulate existing dams. Next, we made clear occurrence and
propagation of cracks in dams during large-scale earthquakes by performing nonlinear dynamic
analysis of a model dam based on the largest class of concrete gravity dams in Japan. Finally, based
on the knowledge gained in this way, we performed a nonlinear dynamic analysis while varying the
acceleration level of the input seismic motion, and we proposed the residual factor of ligament
length in relation to the crack propagation length as a new measure for quantitative evaluation of
seismic safety against crack penetration failure (METI and JEPOC (2001)). Methods for analyzing
the occurrence and propagation of cracks can be broadly divided into discrete crack model and
smeared crack model analysis methods (42).
The seismic behavior of concrete dams has been the subject of extensive research during the
past decade concerning dam safety during earthquakes. Chopra et al (1972), studies seismic
behavior of dam’s crack path by using linear elastic analysis. The analysis shows, places that are in
damage or and risk of the concerning stability of structure. Pal (1976) was the first researcher who
examined Koyna dam by using non-linear analysis. In this research, assuming no effect of reservoir,
being rigid foundation, smeared crack model use for crack expansion and strength criteria to crack
growth, Koyna dam was analyzed and was shown that the results of material properties and
element size are very sensitive. Figure 1(a) crack zone in the dam of which resulting from this
analysis are shown. Skrikerud (1986) studied concrete dams through a case study on Koyna dam and
by employing discrete crack for crack growth and strength criteria for crack expansion. In their
study the growth of crack at each step of growth, the length of the crack tip element was considered
that this is the final results were effective. He interpreted the results of their model, due to
expansion mismatch with the cracking in their analysis of real crack in the dam, no match
Foundation and reservoir interaction and lack of real values of characteristic parameters dam
43
announced. Crack profiles from the analysis left in Figure 1(b) are presented. El-Aidi and Hall
(1989) did a research on seismic fracture of Pine flat dam. Smeared crack model and strength
criteria for crack expansion and growth were used. In their analysis is considering the reservoir –
dam and foundation – dam interaction was crack profile presented. Figure 1(c), cracking in the dam
will provide analysis. Fenves and Vargas-Loli also studied Pine flat dam by using fracture
mechanics criteria for crack growth and smeared crack model to crack expand (Uang and Bertero,
1990). They apply different coefficients of Taft earthquake record, regardless of the effect by
foundation; Pine Flat dam in two cases with and without the effect of the reservoir was analyzed. In
this study the effect of hydrodynamic pressure on the seismic behavior in the dam with the crack
profiles presented. The results of this analysis were shown in Figure 1(d). Bhattacharjee and Leger
(1994) presented detailed results of their study on nonlinear analysis models of gravity concrete
dams. Nonlinear Behavior in Seismic their smeared crack models were used. In their analysis,
dynamic and static results of this model have been compared to laboratory results and cracks formed
in Koyna gravity dam and approved. Results from this analysis in Figure 1(f) can be seen.
Ghaemian and Ghobarah (1999) to assess the seismic behavior of concrete dams in the two-
dimensional space of the smeared crack model presented in 1993 and 1994 by leger and
Bhattacharjee used. 2D seismic fracture behavior of the Koyna dam were examined by Ganglia et al,
(2000).In their studies smeared crack model for crack growth and non-linear fracture mechanics
criteria for the expansion of cracks were used. Lohrasbi and Attarnejad (2008) also used fracture
mechanics theory in their analysis, to evaluate the smeared crack model and discrete crack model.
Nonlinear seismic behavior of the Koyna dam by considering Dam and Reservoir interaction
examined by Mirzabozorg and Ghaemian (2005). They analysis Koyna dam under dynamic loading
defined fracture mechanics of the fracture region and the smeared crack model. Through an article
published in 2005, Calayir and Karaton (2005) evaluated the response of seismic fracture in gravity
concrete dams. In their own study with regard to dam and reservoir interaction, Koyna dam to the
finite element method with Lagrange - Lagrange formula (43).
9.9. Thermal Stress Analysis of Concrete Gravity Dams

Volumetric changes caused by thermal expansion and contraction, or by
alkali/aggregate reactivity affect the cross valley stresses in the dam. These stresses are
important when 3 dimensional behavior is being considered. Expansion will cause a dam
to wedge itself into the valley walls more tightly, increasing its stability. Contraction has
the opposite effect. While these effects are acknowledged, the beneficial effect of
expansion is difficult to quantify even with very elaborate finite element models because
it is contingent on the modulus of deformation of the abutments which is highly variable.
For this reason, the beneficial effects of expansion should not be relied upon in three
dimensional stability analysis. If it appears that contraction will cause monolith joints to
open, and thus compromise force transfer from monolith to monolith, this effect should
be considered (44).
9.10. Relevant Case Histories

Austin (Bayless) Dam: 1911
Austin Dam was a concrete gravity dam about 43 feet high and 534 feet long constructed by
the Bayless Pulp and Paper Company about 1½ miles upstream of the town of Austin, Pennsylvania.
A four-foot-thick by four-foot-deep concrete shear key was constructed into the horizontally bedded
sandstone with interbedded weak shale layers. Anchor bars were grouted 5 to 8 feet into the
foundation, extending well up into the dam body, on 2-foot 8-inch centers, located at about 6 feet
from the upstream face. No drains were provided for the dam or foundation. During initial reservoir
44
filling in 1910, the center portion of the dam at the overflow spillway section slid downstream about
18 inches at the base and 31 inches at the crest. The reservoir was lowered, but no repairs were
made and the dam was put back into service. As the reservoir filled again, the dam suddenly gave
way on September 30, 1911. More than 75 people lost their lives in Austin. Back analysis suggests
that sliding occurred on a weak shale layer within the foundation (41).
Bouzey Dam: 1895
Bouzey Dam was a 72-foot high masonry gravity dam constructed across the L’Aviere River
near Epinal, France. Similar to Austin Dam, the dam was founded on horizontally interbedded
sandstone and lenticular clay seams, with no drainage provisions, and about a 6-foot wide by 10-
foot deep cutoff key constructed into the rock at the upstream face of the dam. Also similar to
Austin Dam, an incident occurred during initial filling whereby the center section of the dam moved
downstream about a foot, shearing the key. Unlike Austin Dam, the reservoir was lowered and the
lower portion of the dam was 20-10 strengthened. Unfortunately, the upper portion of the dam was
quite thin (less than 18 feet thick for about the upper 35 feet), and upon refilling, the dam cracked
and the upper 30 feet or so was sheared off and swept away. Stability calculations indicate that
cracking was likely at the elevation where the shear failure occurred, and once cracked through, the
upper portion of the dam was unstable. (41).
Koyna Dam: 1967
Koyna Dam is a 338-foot-high and 2,800-foot long concrete gravity dam constructed on the
Koyna River in southwestern India between 1954 and 1963. During construction the decision was
made to raise the dam and the downstream slope of the non-overflow section was steepened in the
upper 120 feet of the structure to accommodate the raise, resulting in a discontinuous change in
slope at that location. The dam was shaken by a M6.5 earthquake on December 11, 1967. A strong
motion accelerograph located in a gallery on the upper right abutment recorded a peak ground
acceleration of 0.63g cross-canyon, 0.49g downstream, and 0.34g vertical. Although the dam did
not fail, deep horizontal cracks formed throughout the upstream and downstream faces near the
change in slope where a stress concentration is expected to occur, requiring the installation of
tendons and construction of buttresses on the downstream face to stabilize the structure. Finite
element analyses indicated stress concentrations near the change in slope that exceed the dynamic
tensile strength of the concrete (41).
10. Conclusions
Most rivers have already been managed in some form; not many qualify as pristine. The
World Commission on Dams has reported on the effects of large dams, and small, after their two
year study. In view of the planned technical, financial, and economic performance versus the actual
outcomes, the WCD found large dams to fall short of meeting any of the expected goals (2000). The
ecosystems, biodiversity and downstream livelihoods are well documented and were found to be
degraded, along with limited success in predicting and avoiding loss in these areas (WCD, 2000).
For example, the Commission found ―the use of fish passes to mitigate the blockage of migratory
fish has had little success, as the technology has often not been tailored to specific sites and species‖
(WCD, 2000). Most facilities do not even have fish passages (Marmulla, 2001). Most relocatees are
not adequately provided for either. Knowing one way of life and being asked to move where the
population is denser, or the land is lower quality, or acquiring skills are necessary for a new career,
can be a shock many may not be able to overcome. For the engineers, they have been asked to build
a dam that will enable people to have more food, possibly year round, as well as electricity, drinking
water, and recreation. Those who are asked to sacrifice their homes may get overlooked for a
presumed great good. Declines in fish populations to burying homes and cultures, other methods
45
should be explored in greater depth. The designers of the Three Gorges Project (TGP) were advised
to go the route of several smaller dams, but chose to move forward with the plan in hand. On the
TGP website, several pages boast of being the largest in a number of categories. Many years prior,
notlong after the Hoover Dam was completed, Francis Crowe, ―the [Bureau of Reclamation]
surveyor on the project, later put it, ―I was wild to build this dam—the biggest dam built by anyone,
anywhere‖ .
REFERENCES
1. Golze, A. R. 1977 Handbook of dam engineering. Van Nostrand Reinhold: New York 793 p.
2. Björn Boberg And David Holm (2012) ―FEM Modeling Of Concrete Gravity Dams‖ Degree
Project in Civil Engineering and Urban Management Department of Land and Water Resources
Engineering, Royal Institute of Technology (KTH)
3. T Subramani, D. Ponnuvel (2012) ―Seismic and Stability Analysis of Gravity Dams Using
Staad PRO‖, International Journal of Engineering Research and Development ISSN: 2278-
067X, Volume 1, Issue 5 (June 2012), PP.44-54 44
4. USBR (Ver. No.19) ―Design Criteria For Concrete Arch And Gravity Dams‖.
5. Chinesse COLD (2014) ―Seismic Safety of Concrete Gravity Dams‖.
6. Zhang, C., Song, C., Jin, F., Pekau, O.A., Wang, G., Zhao, C., Xu, Y., Clough, R.W., Yan, C.,
Sun, L., Feng, L., Wang, S., Dong, Y., Li, Q., Ren, Y. & Liu, H. 2001 Numerical modeling of
concrete dam-foundation-reservoir systems. Tsinghua University Press: Beijing 464 p.
7. Zienkiewicz, 0.C. and Taylor, R.L. (2000) “The Finite Element Method”; 5th Edition McGraw-
Hill.
8. Zeidan, B. A. (2014) "Seismic Analysis of Dam-Reservoir-Foundation Interaction for Concrete
Gravity Dams", ", International Symposium on Dams in Environmental Global Challenges"
ICOLD2014, Bali, Indonesia, June 1ST - 6TH, 2014.
9. Westergard, H. M. (1933). Water pressure on dams during earthquakes. TRANSACTIONS
ASCE Vol.98.
10. Chopra A.K. (1967). Hydrodynamic Pressure on dams during earthquakes‖Proc .ASCE , EM6.
11. Chopra A.K . (1970). Earthquake response Analysis of concrete gravity dams. Proc. ASCE,
EM4.
12. Dasgupta G.and Chopra A.K., (1979). Dynamic Stiffness Matrices for Homogeneous
Viscoelastic Half Plane. ASCE, J. of Eng. Mechanics, Vol. 105, pp: 729-745.
13. Chopra A.K. and Chakrabarti P. (1981). Earthquake Analysis of Concrete Gravity Dams
including Dam-Water-Foundation Rock Interaction. Earth. Eng. & Struc. Dynamics, Vol. 9, pp
363-383.
14. Zienkiewicz, 0.C. and Taylor, R.L. (1991). The finite element method; volume II. Forth
edition, First published in I967 by McGraw-Hill pp.407- 419.
15. AKKÖSE, M. A. DUMANO_GLU, A. M. TUNA, E. (2004). Investigation of Hydrodynamic
Effects on Linear and Nonlinear Earthquake Responses of Arch Dams by the Lagrangian
Approach” Turkish J. Eng. Env. Sci. 28 (2004), 25 – 40.
16. Aznarez, J.J. Maeso, 0. and Dominguez, J. (2006). BE analysis of bottom sediments in
dynamic fluid-structure interaction problems. Engineering Analysis with Boundary Elements
30, pp. I24-I36.
46
17. Du, X., Zhang, Y. and Zhang, B. (2007) “ Nonlinear seismic response analysis of arch dam-
foundation systems part I dam-foundation rock interaction‖ Bull Earthquake Eng. (2007)
5:105–119.
18. Akkose, M., Bayraktar, A. and Dumanoglu, A.A. (2008. Reservoir water level effects on
nonlinear dynamic response of arch dams‖ Journal of Fluids and Structures 24 (2008) 418–
435.
19. Seghir, A., Tahakourt, A. and Bonnet, G. (2009). Coupling FEM and symmetric BEM for
dynamic interaction of dam–reservoir systems‖ Engineering Analysis with Boundary Elements
33 (2009) 1201–1210.
20. Bonnet, G., Seghir, A. and Corfdir, A. (2009). Coupling Bem with FEM by a direct
Computation of the boundary stiffness matrix. Comput. Methods Appl. Mech. Eng. I98,
pp.2439-2445.
21. Shariatmadar, H. and Mirhaj, A. (2009). Modal Response of Dam-Reservoir-Foundation
Interaction‖ 8th International Congress on Civil Engineering, May 11-13, 2009, Shiraz
University, Shiraz, Iran.
22. Seleemah, A. A. and El-Sharkawy, M. (2011) .Seismic analysis of base isolated liquid storage
ground tanks., Ain Shams Engineering Journal, Elsevier, 2, pp 33-42.
23. Akhaveissy, A.H. and Malekshahi, M. (2012). Transient Analysis of Dam-Reservoir
Interaction‖ 2012 IACSIT Coimbatore Conferences IPCSIT vol. 28 (2012) © (2012) IACSIT
Press, Singapore.
24. Hamidian, D., Seyedpoor, S.M. and Salajegheh, J. (2013). An Investigation of Dam-Water-
Foundation Rock Interaction Effects on Linear And Nonlinear Earthquake Response of
Concrete Arch Dams‖ Asian Journal Of Civil Engineering (BHRC) VOL. 14, NO.1 pp:111-122
25. Zeidan, B. A. (2013) ―Seismic Dam-Reservoir Interaction of Concrete Gravity Dams‖ 9th
Symposium of ICOLD European Club IECS2013, 10-12 April, Italy.
26. Fenves G. & Chopra A.K. (1984) ―Earthquake Analysis of Concrete Gravity Dams Including
Reservoir Bottom Absorption and Dam- Water-Foundation Rock Interaction”, Earthquake
Engineering and Structural Dynamics, 12:5, 663-680.
27. Fenves, G., And Chopra, A. K., (1985) ―Effects Of Reservoir Bottom Absorption And Dam-
Water-Foundation Rock Interaction On Frequency Response Functions For Concrete Gravity
Dams” Earthquake Engineering & Structural Dynamics, Vol. 13, 1985, Pp. 13-31.
28. Gaun F., Moore I.D. & Lin G. (1994) “Seismic Analysis of Reservoir-Dam-Soil Systems in the
Time Domain”, The 8th international conference on Computer Methods and Advances in
Geomechanics, Siriwardane & Zaman (Eds), Vol. 2, 917-922.
29. Ghaemian M., Noorzad A. & Moghaddam R.M. (2005) “Foundation Effect on Seismic
Response of Arch Dams Including Dam-Reservoir Interaction”, Europe Earthquake
Engineering, 3, 49-57.
30. US. Army Corps of Engineers (USACE), (2003) “Time-History Dynamic Analysis of Concrete
Hydraulic Structures;‖ Chapter 2- Analytical Modeling of Concrete Hydraulic Structures,
Chapter 3-Time-History Numerical Solution Techniques‖, EM 1110-2-6051.
31. Lysmer J. & Kuhlemeyer R.L. (1969) “Finite Dynamic Model for Infinite Media”, Journal of
Engineering Mechanics Division, ASCE, 95 (EM4), 859-877.
32. Wilson E.L. (2000) “Three Dimensional Static and Dynamic Analysis of Structures, A
Physical Approach with Emphasis on Earthquake Engineering”, 4th Ed., Computers and
Structures Inc.
33. Wolf J. P. (1985) “Dynamic Soil-Structure Interaction”, Prentice Hall: Englewood Cliffs, NJ.
34. Bakenaz A. Zeidan (2014) “Finite Element Modeling For Acoustic Reservoir-Dam-
Foundation Coupled System”, International Symposium on Dams in a Global Environmental
Challenges, ICOLD2014, Bali, Indonesia, 1-6 June, 2014.
35. Chopra, A.K., (2012) “Dynamics of Structures: Theory and Applications to Earthquake
Engineering” 7th Edition, Prentice Hall.
36. Zeidan, B. A. (2014) ―Finite Element Modeling For Acoustic Reservoir-Dam-Foundation
Coupled System", International Symposium on Dams in Environmental Global Challenges"
ICOLD2014, Bali, Indonesia, June 1ST - 6TH, 2014.
47
37. Arnkjell Løkke (2013) ―Earthquake Analysis of Concrete Gravity Dams‖ MsC. in Civil and
Environmental Engineering , Norwegian University of Science and Technology.
38. http://www.diva-portal.org/smash/get/diva2:666640/FULLTEXT01.pdf
39. USBR (2012) ―Risk Analysis for Concrete Gravity Structures”
40. Daniel Celarec (2002) “ Probabilistic Seismic Risk Assessment Of Concrete Gravity Dams”
41. http://www.usbr.gov/ssle/damsafety/Risk/BestPractices/20-
ConcreteGravityStructures20120921.pdf
42. H. KIMATA (2008) “SEISMIC SAFETY OF CONCRETE GRAVITY DAMS BASED ON
DYNAMIC CRACK PROPAGATION ANALYSIS DURING LARGE-SCALE EARTHQUAKES “
The 14 World Conference on Earthquake Engineering- October 12-17, 2008, Beijing, China-
43. “Fracture analysis of concrete gravity dam under earthquake induced loads”
44. . CHAPTER III-GRAVITY DAMS-(Revised October 2002)
48
RELATED REFERNCES
 Chopra A.K . (1970). Earthquake response Analysis of concrete gravity dams. Proc. ASCE, EM4.
 Du, X., Zhang, Y. and Zhang, B. (2007) ― Nonlinear seismic response analysis of arch dam-
foundation systems part I dam-foundation rock interaction‖ Bull Earthquake Eng. (2007) 5:105–119.
 P. Novak, A.I.B. Moffat , C. Nalluri and R. Narayanan (2007) ―Hydraulic Structures‖ Forth
Edition, Taylor & Francis Co.
 Heidi Hull (2009) ‖Large Dams- Human and Environmental Benefits and Costs‖ University Of
Wisconsin‐Stevens Point.
 Dasgupta G.and Chopra A.K., (1979). Dynamic Stiffness Matrices for Homogeneous Viscoelastic
Half Plane. ASCE, J. of Eng. Mechanics, Vol. 105, pp: 729-745.
 Chopra A.K. and Chakrabarti P. (1981). Earthquake Analysis of Concrete Gravity Dams including
Dam-Water-Foundation Rock Interaction. Earth. Eng. & Struc. Dynamics, Vol. 9, pp 363-383.
 Zienkiewicz, 0.C. and Taylor, R.L. (1991). The finite element method; volume II. Forth edition,
First published in I967 by McGraw-Hill pp.407- 419.
 AKKÖSE, M. A. DUMANO_GLU, A. M. TUNA, E. (2004). Investigation of Hydrodynamic

Effects on Linear and Nonlinear Earthquake Responses of Arch Dams by the Lagrangian Approach‖
Turkish J. Eng. Env. Sci. 28 (2004), 25 – 40.
 Aznarez, J.J. Maeso, 0. and Dominguez, J. (2006). BE analysis of bottom sediments in dynamic
fluid-structure interaction problems. Engineering Analysis with Boundary Elements 30, pp. I24-I36.
 Daniel Celarec (2002) “ Probabilistic Seismic Risk Assessment Of Concrete Gravity Dams‖
 Akkose, M., Bayraktar, A. and Dumanoglu, A.A. (2008. Reservoir water level effects on nonlinear
dynamic response of arch dams‖ Journal of Fluids and Structures 24 (2008) 418–435.
 Seghir, A., Tahakourt, A. and Bonnet, G. (2009). Coupling FEM and symmetric BEM for dynamic
interaction of dam–reservoir systems‖ Engineering Analysis with Boundary Elements 33 (2009)
1201–1210.
 Bonnet, G., Seghir, A. and Corfdir, A. (2009). Coupling Bem with FEM by a direct Computation of
the boundary stiffness matrix. Comput. Methods Appl. Mech. Eng. I98, pp.2439-2445.
 Shariatmadar, H. and Mirhaj, A. (2009). Modal Response of Dam-Reservoir-Foundation

Interaction‖ 8th International Congress on Civil Engineering, May 11-13, 2009, Shiraz University,
Shiraz, Iran.
 Seleemah, A. A. and El-Sharkawy, M. (2011) .Seismic analysis of base isolated liquid storage
ground tanks., Ain Shams Engineering Journal, Elsevier, 2, pp 33-42.
 Elkhoraby, S., Seleemah, A. A. and El-Sharkawy, M. (2011). Validity of simplified models for
prediction of the seismic response of isolated liquid storage ground tanks., the 2011 World
Congress on Advances in Structural Engineering and Mechanics (ASEM'11+), Seoul, Korea, 18-22
September, 2011, pp 3279-3292.
49
 Seleemah, A. A. and El-Sharkawy, M. (2011). Seismic analysis and modeling of isolated elevated
liquid storage tanks., Earthquakes and Structures, An Int. Journal, Vol. 2 No. 4., Dec. 2011, pp 397-
412.
 Akhaveissy, A.H. and Malekshahi, M. (2012). Transient Analysis of Dam-Reservoir Interaction‖

2012 IACSIT Coimbatore Conferences IPCSIT vol. 28 (2012) © (2012) IACSIT Press, Singapore.
 Hamidian, D., Seyedpoor, S.M. and Salajegheh, J. (2013). An Investigation of Dam-Water-

Foundation Rock Interaction Effects on Linear And Nonlinear Earthquake Response of Concrete
Arch Dams‖ Asian Journal Of Civil Engineering (BHRC) VOL. 14, NO.1 pp:111-122
 El-Sharkawy, M. E. M. (2008). Seismic Analysis of Seismically Isolated Water Tanks. MSc. Thesis,
Structural Engineering Department, Faculty of Engineering, Tanta University, Egypt.
 Sommerfeld A. (1949) ―Partial Differential Equations in Physics‖, Academic Press, New York.
 Hall, J. F., and Chopra, A. K., (1982) ―Hydrodynamic Effects in the Dynamic Response of Concrete
Gravity Dams”, Earthquake Engineering and Structural Dynamics, Vol. 10, No 2, 1982, pp. 333-
345.
 Millan, M. A., Young, Y. L. and Prevost, J. H. (2007) ― The Effect of Reservoir Geometry on the
Seismic Response of Gravity Dams‖, Earthquake Eng. Struct. Dyn. 2007; 36:1441–1459.
 Chopra, A.K., (2012) ―Dynamics of Structures: Theory and Applications to Earthquake

Engineering‖ 7th Edition, Prentice Hall.
 Bakenaz A. Zeidan (2013) “Hydrodynamic Analysis of Concrete Gravity Dams Subjected To

Ground Motion‖ 9th Symposium of ICOLD European Club. Club IECS201310-12 April, Italy.
 Fenves G. & Chopra A.K. (1984) ―Earthquake Analysis of Concrete Gravity Dams Including
Reservoir Bottom Absorption and Dam- Water-Foundation Rock Interaction‖, Earthquake
Engineering and Structural Dynamics, 12:5, 663-680.
 Fenves, G., And Chopra, A. K., (1985) ―Effects Of Reservoir Bottom Absorption And Dam-Water-
Foundation Rock Interaction On Frequency Response Functions For Concrete Gravity Dams‖
Earthquake Engineering & Structural Dynamics, Vol. 13, 1985, Pp. 13-31.
 Gaun F., Moore I.D. & Lin G. (1994) “Seismic Analysis of Reservoir-Dam-Soil Systems in the Time
Domain”, The 8th international conference on Computer Methods and Advances in Geomechanics,
Siriwardane & Zaman (Eds), Vol. 2, 917-922.
 Ghaemian M., Noorzad A. & Moghaddam R.M. (2005) “Foundation Effect on Seismic Response of
Arch Dams Including Dam-Reservoir Interaction”, Europe Earthquake Engineering, 3, 49-57.
 US. Army Corps of Engineers (USACE), (2003) “Time-History Dynamic Analysis of Concrete
Hydraulic Structures;‖ Chapter 2- Analytical Modeling of Concrete Hydraulic Structures, Chapter
3-Time-History Numerical Solution Techniques‖, EM 1110-2-6051.
 Lysmer J. & Kuhlemeyer R.L. (1969) “Finite Dynamic Model for Infinite Media”, Journal of
Engineering Mechanics Division, ASCE, 95 (EM4), 859-877.
 Wilson E.L. (2000) ―Three Dimensional Static and Dynamic Analysis of Structures, A Physical
Approach with Emphasis on Earthquake Engineering‖, 4th Ed., Computers and Structures Inc.
 Wolf J. P. (1985) “Dynamic Soil-Structure Interaction”, Prentice Hall: Englewood Cliffs, NJ.
50
 Zeidan, B. A. (2013) ―Seismic Dam-Reservoir Interaction of Concrete Gravity Dams‖ 9th

Symposium of ICOLD European Club IECS2013, 10-12 April, Italy.
 Bakenaz A. Zeidan (2014) “Finite Element Modeling For Acoustic Reservoir-Dam-Foundation

Coupled System”, International Symposium on Dams in a Global Environmental Challenges,
ICOLD2014, Bali, Indonesia, 1-6 June, 2014.
 Ministry of Water Recourses and Electric Power of People’s Republic of China. 1979a. Design
Specification for Concrete Gravity Dams (SDJ 21-78). China Water Press: Beijing.
 Ministry of Water Recourses and Electric Power of People’s Republic of China. 1979b.
Supplementary provisions of Design specification for Concrete Gravity Dams (SDJ 21-78). China
Water Press: Beijing.
 Arnkjell Løkke (2013) ―Earthquake Analysis of Concrete Gravity Dams‖ MsC. Thesis, ,Civil and
Environmental Engineering, Norwegian University of Science and Technology.
 Amadei, B., T. Illangasekare, C. Chinnaswamy, D.I. Morris, ―Estimating Uplift in Cracks in

Concrete Dams,‖ Proceedings, International Conference on Hydropower, Denver, Colorado, July
24-26, 1991.
 Anderson, C., C. Mohorovic, L. Mogck, B. Cohen, G. Scott, ―Concrete Dams Case Histories of
Failures and Nonfailures with Back Calculations,‖ Report DSO-98-005, Bureau of Reclamation,
Denver, Colorado, December 1998.
 Cannon, R.W., ―Appendix E, Tensile Strength of Roller Compacted Concrete,‖ EP 1110-2-12, U.S.
Army Corps of Engineers, September 30, 1995.
 Bureau of Reclamation, Design of Small Dams, Third Edition, Denver, CO, 1987.
 Hendron, A.J., Jr., E.J. Cording, and A.K. Aiyer, ―Analytical and Graphical Methods for the
Analysis of Slopes in Rock Masses,‖ Technical Report GL-80-2, prepared for 20-11
 U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi, March 1980.
 Raphael, J.M., ―Tensile Strength of Concrete,‖ Title No. 81-17, ACI Journal, March-April, 1984,
pp. 158-165.
 Watermeyer, C.F., ―A Review of the Classical Method of Design of Medium Height Gravity Dams
and Aspects of Base Shortening with Uplift,‖ Journal of the South African Institution of Civil
Engineering, Vol 48 No 3, pp. 2-11, 2006.
 Mills-Bria, B.L., L.K. Nuss, D. Harris, D.H.O’Connell, ―State-of-Practice for Non-Linear Analysis
at the Bureau of Reclamation.‖
 Adedeji, A. A. (2004); Finite Element Method, CVE 567 Lecture Notes, Department of Civil
ngineering, University of Ilorin, Ilorin.
 Bathe, K. J. (1996); Finite Element Procedures, Seventh Edition, PrenticeHall Inc, New Jersey.
 CIWAT ENGINEERS (1996); Design Report of the University of Ilorin Dam, CIWAT
ENGINEERS, Ilorin, pp 14 – 25.
 EM 111022200 (1995); Design of Concrete Gravity Dams, Engineer Manual of the US Army Corps
of Engineers, US Army Corps Publications Depot, National Technical Information Service,
Springfield, VA, www.us.armycorps.engineers.com/engineer’s manual.
51
 Fenves, G. L. and VargasLoli, M. (1988); NonLinear Analysis of FluidStructure Systems, Journal

Of Engineering Mechanics Division, ASCE Vol. 114, pp 219 – 240.
 Hatami, K. (2001); Seismic Analysis of Concrete Dams, National Defence, Royal Military College
of Canada, www.zworks.com/seismic analysis/concrete dams/ Seismic Analysis of Concrete Dams.
 Iroko, A. O. (2001); Hysterical Analysis of Strawbale as an Infill Material Subjected to Seismic

Loadings (Vibration), B. Eng. Project, Submitted to the Dept. of Civil Engineering, University of
Ilorin, Ilorin.
 Lotfi, V. (2001); Seismic Analysis of Concrete Gravity Dams using Decoupled Modal Approach in
Time Domain, Electronic Journal Of Structural Engineering, Vol. 3, www.ejse.org, pp 102 – 116.
 Major, A. (1980); Dynamics in Civil Engineering, Vol. I IV, Second Edition, Collet’sHoldings Ltd,
London.
 Polyakov, S. V. (1985); Design of EarthquakeResistant Structures, Second Edition, MirPublishers.
Related Cited:
 http://www.usbr.gov/ssle/damsafety/Risk/BestPractices/20-
ConcreteGravityStructures20120921.pdf
 Agostinho, C.S., Hahn, N.S., & Marques, E.E. (2003). Patterns of food resource used by two
congeneric species of piranhas. Brazilian Journal of Biology. 63(2). pp177-182. Abstract.
Retrieved May 2, 2009 from Web Site:
http://csaweb116v.csa.com/ids70/view_record.php?id=1&recnum=0&SID=8e1u9k26be1do3bnoeff2
eq5s6&mark_i d=search%3A1%3A0%2C0%2C1
 Atkins, W.A. (2007). Colorado River Basin. In Water Encyclopedia: Science and Issues.
Retrieved April 4, 2009, from Web Site: http://www.waterencyclopedia.com/Ce-Cr/Colorado-River-
Basin.html
 Barlow, M., & Clarke, T. (2002). Blue Gold: The Fight to Stop the Corporate Theft of the
World's Water. New York City: The New Press.
 Bily, C.A. (2000). Hetch Hetchy Dam. In Encyclopedia of Environmental Issues Volume II.
Pasadena, CA; Salem Press, Inc.
 BDS (2009). About Dams. Retrieved February 28, 2009 from British Dam Society Web site:
http://www.britishdams.org/about_dams/types.htm
 Carmichael, R.S. (2000). Three Gorges Dam. In Encyclopedia of Environmental Issues Volume
III. Pasadena, CA: Salem Press, Inc.
 CIA (2009) The World Fact Book. Retrieved April 1, 2009, from the Central Intelligence Agency
Web Site:
https://www.cia.gov/library/publications/the-world-factbook/index.html
 Clarke, R., & King, J. (2004). The Water Atlas. New York, NY: The New Press.
 Clarkin, T. (2000). Boulder Dam. In Encyclopedia of Environmental Issues Volume I.
Pasadena, CA: Salem Press, Inc.
 Corfield, J. (2007). Hoover Dam In Encyclopedia of Environment and Society Volume I.
Thousand Oaks, CA: SAGE Publications, Inc.
 CTGP (2002). China Three Gorges Project. Retrieved January 24, 2009 from CTGP Web Site:
http://www.ctgpc.com/index.php
 CTIC (2009) What is a Watershed? Retrieved May 5, 2009 from the Conservation Technology
Information Center Web Site:
http://www.conservationinformation.org/?action=learningcenter_kyw_whatisawatershed
 Davidson, F.P. & Brooke, K.L. (2006). Building the World: An Encyclopedia of the Great
Engineering Projects in History Volume Two. Westport, CT: Greenwood Press.
 Dudgeon, D. (2000). The Ecology of Tropical Asian Rivers and Streams in Relation to Biodiversity
Conservation.
52
 Annual Review of Ecology and Systematics. 31, pp239-263. Retrieved February 15, 2009 from
JSTOR Web Site: http://www.jstor.org/stable/221732
 DWMI (2009). Colorado River Basin: Lifeline of the Southwest. Desert USA. Retrieved April 5,
2009 from Digital West Media, Inc. Web Site:
http://www.desertusa.com/colorado/coloriv/du_coloriv.html
 Echeverria J., Barrow, P., & Roos-Collins R. (1989). Rivers at Risk: A Concerned Citizen's
Guide to Hydropower. Washington D.C.: Island Press.
 Edwards, L.M. & Redmond, K.T. (2005). Climate Factors on Colorado River Basin Water Supply.
Colorado River
 Basin Climate: Paleo, Present, Future Retrieve April 3, 2009 from California Department of Water
Resources Web Site: http://www.water.ca.gov/drought/docs/co_nov05.pdf
 EERE (2008). Wind and Hydropower Technologies Program: History of Hydropower. Retrieved
March 6, 2009, from U.S. Department of Energy, Energy Efficiency and Renewable Energy Web
site:
http://www1.eere.energy.gov/windandhydro/hydro_history.html
 FEMA (2006). Benefits of Dams. Retrieved February 29, 2009, from United States Department of
Homeland
Security Federal Emergency Management System Web site:
http://www.fema.gov/hazard/damfailure/benefits.shtm
 Ford, N. (2007). Facing up to the challenges of the Nile. International Water Power and Dam
Construction. Retrieved March 28, 2009 from IWPC Web Site:
http://www.waterpowermagazine.com/story.asp?storyCode=2046657
 George, A. (1998). Friction flows over Nile waters. Water Power Magazine. Retrieved March 28,
2009 from International Water Power & Dam Construction Web Site:
 Gibbs, W.W. (2002). The Power of Gravity. Scientific American. 287(1) pp88. Retrieved March
28, 2009, from Ebscohost Search at Web Site:
http://web.ebscohost.com.ezproxy.uwsp.edu/ehost/detail?vid=1&hid=108&sid=62e2a5c0-a033-
4d7c-b5b3-
1d3607806e07%40sessionmgr108&bdata=JkF1dGhUeXBlPWlwLGNvb2tpZSx1cmwsdWlkJnNpd
GU9ZWhvc3QtbGl2ZQ%3d%3d
 Goldsmith, E., & Hildyard, N. (1984). The Social and Environmental Effects of Large Dams.
San Francisco, CA: Sierra Club Books. pg. 225
 Heming, et al, et al, L., Waley, P., & Rees, P. (2001). Reservoir resettlement in China: past
experience and the Three Gorges Dam. The Geographical Journal 167( 3) pp195-212. Retrieved
April 10, 2009 from The Geographical Journal Web Site:
http://www.articlearchives.com/environment-natural-resources/land-use/943371-1.html.
 International Water Power (1998). Needle valves replaced at Hoover Dam. Water Power
Magazine. Retrieved March 28, 2009 from IWP&DC Web Site:
 Kich, M. (2007). Aswan High Dam. In Encyclopedia of Environment and Society Volume I.
Thousand Oaks, CA; SAGE Publications, Inc.
 Krauter, S. (1998). Itaipu: Largest Power Plant on Earth, 12,600 MW of Hydro Power. Universidade
Federal do Rio de Janeiro-Lab Fotovoltaico Inhaltsverzeichnis. Retrieved April 4, 2009 from UFRJ
Web Site:http://www.solar.coppe.ufrj.br/itaipu_ee.html
 Lanz, K. (1995). The Greenpeace Book of Water. New York: Sterling Publishing Co., Inc.
 Leslie, J. (2005). Deep Water: The Epic Struggle over Dams, Displaced People, and the
Environment. New York, NY: Farrar, Straus, and Giroux.
 Lopez, M.A., Acevedo, H.A., & Vazquez, F.A. (2000). Water Quality of the Parana River at
Corrientes, Argentina: A Ten Year Record. The Electronic Journal of the International
Association for Environmental Hydrology on the World Wide Web Volume 8. Retrieved April
10, 2009 from JOEH Web Site: http://www.hydroweb.com/jeh/jeh2000/vazq.pdf
 Luna, S.M., Torres, A.G., Jegu, M. (2009). Serrasalmus spiloplueura, Speckled piranha.
Retrieved May 4, 2009 from Fishbase Website:
http://www.fishbase.org/Summary/SpeciesSummary.php?id=11973
53
 Marmulla, G. (2001). Dams, fish and fisheries: Opportunities, challenges and conflict resolution.
FAO Fisheries Technical Paper 419. Retrieved May 2, 2009 from Food and Agriculture
Organization of the United Nations Web Site: ftp://ftp.fao.org/docrep/fao/004/Y2785E/y2785e.pdf
 MSNBC (2007). Behold the incredible shrinking Colorado River: Experts: Even worse water
shortages possible due to warming, population. Retrieved April 4, 2009 from MSNBC Web Site:
http://www.msnbc.msn.com/id/17276693/
 NID (2004). CorpsMap: National Inventory of Dams. Retrieved April 2, 1009 from U.S. Army
Corps of Engineers Web Site: https://rsgis.crrel.usace.army.mil/apex/f?p=397:1:3893779246060456
 NPS (2009). Lake Mead: Lake Levels and Temperatures. Retrieved April 5, 2009, from the National
Parks Service Web Site: http://www.nps.gov/lame/planyourvisit/weather.htm
 Parker, S.P. (ed.) (1995a). Mediterranean Africa: Egypt. In World Geographical Encyclopedia
Volume 1. New York, NY: McGraw-Hill, Inc.
 Parker, S.P. (ed.) (1995b). The Americas: Anglo-Saxon America. In World Geographical
Encyclopedia Volume 2. New York, NY: McGraw-Hill, Inc.
 Parker, S.P. (ed.) (1995c). Asia. In World Geographical Encyclopedia Volume 3. New York,
NY:McGraw-Hill, Inc.
 Partlow, J. (2008). Doubt, Anger Over Brazil Dams. The Washington Post. Retrieved April 2, 2009
from The Washington Post Web Site
http://www.washingtonpost/wp-dyn/content/article/2008/10/13/AP2008101302539.html
 PBS (2001). Dams. Building Big. Retrieved January 24, 2009, from PBS Web Site:
http://www.pbs.org/wgbh/buildingbig/dam/index.html
 Pearce, F. (2006). When the Rivers Run Dry: Water--the Defining Crisis of the Twenty-First
Century. Boston, MA: Beacon Press.
 Peterson, E.T. (1954). Big Dam Foolishness: The Problem of Modern Flood Control and
Water Storage. New York, NY: The Devin-Adair Company.
 Pitzl, G.R. (2007). Colorado River. In Encyclopedia of Environment and Society Volume I.
Thousand Oaks, CA: SAGE Publications, Inc.
 Polaha, M. & Ingraffea, A.R. (1999). Types of Dams. The Cracking Dams. Retrieved February 23,
2009, from
 SimScience Web site: http://simscience.org/cracks/advanced/mintro.html
 Richter, B., & Thomas, G.A. (2008). Dam good operations. Water Power Magazine. Retrieved
March 3, 2009 from International Water Power & Dam Construction Web Site:
 http://www.waterpowermagazine.com/story.asp?storyCode=2050148
 Rothfeder, J. (2001). Every Drop for Sale: Our Desperate Battle over Water in a World About
to Run Out. New York, NY: Penguin Putnam.
 Rotuli, M. (2008). Hydropower. U.S. Bureau of Reclamation, Department of the Interior. Retrieved
June 29, 2009 from USBR Web Site: http://www.usbr.gov/power/data/hydbroch.pdf
 Salisbury, N.E. (2000). Flood Control. In Encyclopedia of Environmental Issues Volume I.
 Skinner, J. (2000). Sharing Dam Experiences. Retrieved March 28, 2009, from International Water
Power and Dam Construction Magazine Web Site:
 Thompson, D.J (2000). Aswan High Dam. In Encyclopedia of Environmental Issues Volume I.
 UNESCO-WWAP (2003). Water for People Water for Life. Barcelona, Spain: United Nations
Educational, Scientific, and Cultural Organization & Berghahn Books.
 United Nations (2002). Global Challenge, Global Opportunity: Trends in Sustainable Development.
Retrieved June 29, 2009 from the United Nations Web Site:
http://www.un.org/esa/sustdev/publications/critical_trends_report_2002.pdf
 U.S. ACE (2007). A Brief History: Multipurpose Waterway Development. Retrieved March 5, 2009
from U.S.
Army Corps of Engineers Web site: http://www.usace.army.mil/History/Pages/Brief/07-
development/develop.html
 USBR (2004a). Hoover Dam: Fortune Magazine September 1933. Retrieved March 28, 2009, from
Reclamation Managing Water in the West: Lower Colorado Region Web site:
54
http://www.usbr.gov/lc/hooverdam/History/articles/fortune1933.html
 USBR (2004b). Hoover Dam: Workforce. Retrieved March 28, 2009, from Reclamation Managing
Water in the West: Lower Colorado Region Web site:
http://www.usbr.gov/lc/hooverdam/History/essays/workforc.html
 USB R (2008). Hoover Dam Frequently Asked Questions and Answers: The Dam. Retrieved March
28, 2009, from Reclamation Managing Water in the West: Lower Colorado Region Web site:
http://www.usbr.gov/lc/hooverdam/faqs/damfaqs.html
 USGS (2008). Hydroelectric Power: How it Works. Water Science for Schools. Retrieved February
28, 2009 from U.S. Geological Survey Web site: http://ga.water.usgs.gov/edu/hyhowworks.html
 WCD (2000). Dams and Development: A New Framework for Decision-Making. Retrieved
February 29, 2009 from
 World Commission on Dams Website: http://www.dams.org/report/wcd_overview.htm
 WRCC (2006a). Climate of Colorado. Historical Climate Information. Retrieved April 3, 2009, from
http://www.wrcc.dri.edu/narratives/COLORADO.htm
 WRCC (2006b). Average Statewide Precipitation for Western U.S. States. Retrieved April 3, 2009,
from
http://www.wrcc.dri.edu/htmlfiles/avgstate.ppt.html
 Yang, H., Haynes, M., Winzenread, S., & Okada, K. (1999). History of Dams: Examples of Dam
Types, Diagrams. Retrieved February 28, 2009 from University of California Davis Web site:
http://cee.engr.ucdavis.edu/faculty/lund/dams/Dam_History_Page/Diagrams.htm#Gravity%
55
View publication stats

Zeidan Damchapter 2014

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Zeidan Damchapter 2014

Hochgeladen von

Copyright:

Verfügbare Formate

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

Design and Analysis of Concrete Gravity Dams

Research · August 2015

The Nile River Book View project

Isotope hydrology and water resources View project

The user has requested enhancement of the downloaded file.

Dr. Bakenaz A. Zeidan

State of Art in Design and Analysis of

1.1. Implementation of Concrete Gravity Dams

1.2. Design Procedure of Concrete Gravity Dams

1.3. Safety Criteria of Concrete Gravity Dams

most of researchers. When such system is

2. ABOUT CONCRETE GRAVITY DAMS

Figure 1: Different types of concrete dams (2).

 Axis of the dam

means its structural height.

2.2. Dam Concrete Static Properties

2.3. Dam Concrete Dynamic Properties (USBR)

2.4. Foundation Properties

2.4.3. Pore Pressure and Permeability

2.4.5. Compressive and Tensile Strength

3. GRAVITY DAM LOAD COMBINATION

3.1. Static Loads:

3.2. Seismic Loads

where ax, α, g are the ground

3.3. Load Combinations

A usual load case occurs often, or

3.4. Stability criteria

3.5. Factors of Safety

Criteria (USBR) Figure (11)

4.2. Dam-Reservoir Boundary Condition:

4.3. Reservoir-Foundation Boundary Condition:

where is the damping coefficient characterizing the effects of absorption of hydrodynamic

4.4. Reservoir-Far-End Boundary Condition

4.5. Free-Surface Boundary Condition

Figure13: Dam-Reservoir-Foundation System

5. FINITE ELEMENT MODELING

5.1. Strain and Stress in FEM

5.2. FEM Formulation

and the fluid is similarly discretized as

In which represents the nodal ground acceleration vector.

5.3. Fluid-Structure Interaction

Figure 14: Dam-Reservoir coupled system (25).

5.4. Fluid-Structure-Foundation Interaction

researchers. More efficient methods

simple pseudo-static method initially

5.5. FEM Modeling Assumptions

6. ANALYSIS OF CONRETE GRAVITY DAMS

5.1. Gravity Method

5.2. Finite Element Method

6.3. Modal Analysis and Natural Response

6.4. Dynamic Equilibrium

DAM EQUATION OF MOTION (Zeidan 2013)

RESERVOIR EQUATION OF MOTION

6.5. Dynamic Analysis

9. Select the dynamic methods of analysis to be used.

6.5.1. Pseudo Dynamic Method

4.3.2.1 Response Spectrum Method

4.3.2.2 Time History Method

acceleration record used. For this

4.4.3. Direct Solution Method

where α is the ground acceleration, then the resulting pressure is given by :

7. DESIGN & SAFETY CRITERIA