Sinteza articole

-identificare probleme-
1.Combined Seepage and Slope Stability Analysis
of Rapid Drawdown Scenarios for Levee Design
The scope of this paper
-Does an effective stress analysis diverge significantly from the 3-stage Duncan
(1990) analysis?
-If so, under what conditions?
-This paper compares the Duncan (1990) three-stage methodology for analyzing
rapid drawdown scenarios to a combined transient seepage and slope stability
analysis
-Analyses of a number of typical cross-sections will be considered in order to
determine the potential influence of geometry.

The intent of the paper is to illustrate scenarios under which the Duncan (loc. cit.)
methodology produces similar results to the results of a more rigorous analysis
1.Combined Seepage and Slope Stability Analysis
of Rapid Drawdown Scenarios for Levee Design

Rapid draw-down scenario is one of the most severe loading conditions which
can afflict a levee
The pore-water pressures present in the levee during the flooding do not have
enough time to dissipate.
Particularly true for claytype materials where the hydraulic conductivity of the
material is relatively low. A situation is therefore created in which heightened
pore-water pressures on the up-stream side of a levee can trigger either deep
or shallow failures
rapid draw-down scenarios can be analyzed by either a i) total stress or ii) an
effective stress analysis
This paper explores a comparison between the total stress and effective stress
methodologies for typical material types
1.Combined Seepage and Slope Stability Analysis
of Rapid Drawdown Scenarios for Levee Design
free draining soils the design procedure that is recommended is an effective stress
analysis where the initial and final pore-water pressure levels are determined from
a steady-state analysis where the initial and final conditions of the water table are
determined using two separate steady-state seepage analysis.
low permeability soils, the design manual recommends a three-stage approach
which uses a combination of effective strength results and consolidated-undrained
(total) strength results to estimate a worst-case scenario that represents a
conservative design.
three-stage approach represents a methodology based partly on a total stress
analysis as a limiting condition.
three-stage procedure has evolved from first version called the Lowe and Karafiath
(1959) method and later to the USACE (1970) method. Duncan et al. (1990)
reviewed both of these methods and suggested an alternative three-stage analysis
procedure
1.Combined Seepage and Slope Stability Analysis
of Rapid Drawdown Scenarios for Levee Design

The Duncan method limitations

1. The time over which rapid drawdown occurs is not accounted for in the
procedure,
2. The method assumes that a consolidated-undrained laboratory test represents
the limiting condition along the entire critical slip surface. A single value of
undrained shear strength is not appropriate along the entire slip surface (Kerkes, et
al., 2003),
3. The determination of an appropriate value for the undrained shear strength
value for the analysis can be complicated (Kerkes, et al., 2003),
4. The location of the critical slip surface is assumed to be deep and to not change
location during the rapid drawdown sequence.
Duncan et al. (loc. cit.) approach was conceived and designed in a time when a
geotechnical engineers did not have access to software tools required in order to
perform a transient saturated/unsaturated seepage analysis
1.Combined Seepage and Slope Stability Analysis
of Rapid Drawdown Scenarios for Levee Design
1. Benchmark the Duncan method,
2. Pilarcitos comparison,
3. Walter Bouldin comparison.
It should be noted that there are a significant number of input variables which can influence the
outcome of the analysis. Some of the variables include:
Slip surface location (deep / shallow)
Stress state approach (effective stress / total stress)
Saturated / unsaturated shear strength conditions
Variance in seepage or stress - deformation material properties
Variance in the slope angle
Variance in material heterogeneity
Variations in slope stability calculation methodology (i.e., Spencer, Morgenstern-Price, GLE, etc.)
1.Combined Seepage and Slope Stability Analysis
of Rapid Drawdown Scenarios for Levee Design
Conclusion
The analyses presented in this paper show that there are differences between a
total stress and an effective stress analysis.
The primary difference is related to consideration of transient porewater
pressures in the slope stability analysis.
A secondary influence is the effect of geometry on the flow regime in an earth
dam or levee.
The implementation of the rapid drawdown presented by Duncan et al. (1990)
has been benchmarked and the differences are reasonable.
The time over which rapid draw-down occurs can be simulated in the
saturated/unsaturated effective stress transient seepage analyses.
In the case of a multi-region earth dam (such as Walter Bouldin) an effective
stress analysis can allow detailed insight into the mechanism of failure.
2.ANALYSIS OF TRANSIENT EMBANKMENT
STABILITY USING THE DYNAMIC PROGRAMMING
METHOD
Conventional limit equilibrium methods of slope stability analysis are usually
combined with restrictions to the slip surface shape. The Dynamic
Programming Method (DPM) of slope stability analysis represents an
important breakthrough in that these restrictions are significantly relaxed.
This paper presents the analysis of transient embankment stability using a
computer model based on the DPM, combined with finite element stress and
pore-water pressure fields.
The problem studied contains sharp wetting fronts and a significant,
nonlinear spatial variation of shear strength. The analyses results indicate
that the DPM model is robust, fast, and stable. The stability analysis using
the DPM shows that the slip surface shape is affected by infiltration and
deviates from a circular shape in some cases.
Poisson ratio was found to have an influence on the shape of critical slip
surfaces and the factors of safety.
2.ANALYSIS OF TRANSIENT EMBANKMENT
STABILITY USING THE DYNAMIC PROGRAMMING
METHOD
A hypothetical embankment comprised of lacustrine silt subjected
to a continuous, heavy rainfall was analyses for a period of four
days, using a dynamic Programming algorithm called SAFE-DP. The
stresses and pore-water pressures were determined based on the
partial differential equations governing equilibrium and moisture
flow. The general purposed solver called FlexPDE (PDE Solutions Inc.
2003) was used for the finite element analyses.
The results of the analyses indicate that the DPM code is robust,
fast, and stable. The stability analysis using the DPM shows that the
slip surface shape is affected by the infiltration and deviates from a
circular shape in some cases. Poisson ratio was found to have an
important influence on the shape of critical slip surfaces and the
factors of safety.
2.ANALYSIS OF TRANSIENT EMBANKMENT
STABILITY USING THE DYNAMIC PROGRAMMING
METHOD
The DPM is a general method of maximization and minimization of linear (additive)
functionals (Bellman 1957).
The combined use of finite element stress fields further enhances the analysis by
providing means of incorporating into the slope stability analysis more realistic
boundary conditions, soil stressstrain properties, and stress history.
The dynamic programming procedure is used to determine a continuous
assemblage of segments that corresponds to a minimum overall factor of safety.
In order to quantify the overall stability, a factor of safety, Fs, along a slip surface
is defined in its discrete form
The unique characteristic of the DPM lies in the manner by which the shape and
position of the critical slip surface, along with the corresponding minimum Fs, is
obtained. The dynamic programming method can only be applied to additive
functionals (i.e., functions of the form Y = S[ayi + b]). For that reason, in order to
minimize the non-additive Fs functional the following additive functional is
introduced (Baker, 1980):
3.Effect Of Mesh Resolution On Long-Term Water
Balance Calculations

Such numerical models are run with the use of complex climatological
data and unsaturated soil property functions. The input information is
then used by the finite element solvers to determine vertical flow
rates and, ultimately, long-term percolation rates through the earth
cover.
Calculations are generally complicated by the fact that the infiltration
into a dry soil is one of the more complex types of numerical modeling
scenarios. The challenge occurs largely because of the non-linearities
present in the unsaturated soil properties.
This paper examines the numerical difficulties associated with these
calculations and examines the impact of small numerical issues over
long time periods. In particular, the role of mesh density as it is
related to the accuracy of water balance calculations is examined.
3.Effect Of Mesh Resolution On Long-
Term Water Balance Calculations
The mesh density of a numerical model is one of the primary sources
of error (Yeh, 2000). In order to demonstrate this issue a simple
numerical model is set up in the SVFlux software. The numerical
model is a simple fixed-mesh model consisting of a vertical column of
soil with unsaturated soil properties defined. A few random
precipitation events are then applied to the top of the soil column and
the impact of these precipitation events is then tracked in the model.
The soil-water characteristic curve is first then steepened until the
results of the numerical model become erroneous ( Table 1). The
unsaturated hydraulic conductivity was estimated using the Modified
Campbell method. Once the numerical model is brought to failure,
the mesh density in the numerical model is increased until the model
again reaches the correct answer.
3.Effect Of Mesh Resolution On Long-
Term Water Balance Calculations
A second experiment was then set up as a numerical model. In the
numerical model the total modeling time is set to 7 days. In each of these
days a progressively more intense storm is introduced to the numerical
model. The particular storm events are presented in the following table
(Table 2).
The model is: i. first run with automatic mesh refinement turned on. Then
the model is run with a default static and unrefined mesh. The static
unrefined mesh yields large volume-mass errors in the calculations. The
mesh density is then increased in the staticmesh scenario until a static
mesh yields a theoretically correct answer.
It can be seen with this study that the numerical modeling time required
by numerical models performing cover modeling has the potential to be
reduced through the selective use of additional nodes only in times of
peak intensity storms
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis
A practical three-dimensional slope stability approach is
presented.
Most slope stability problems are three-dimensional in
nature. Several field conditions can be better represented
by three-dimensional models, such as excavation fronts,
slope corners, dam shoulders.
This paper presents how three-dimensional slope stability
analyses can be undertaken using simple finite element
stress and seepage analysis. Two benchmark problems are
presented in order to demonstrate the accuracy of the
method of analysis.
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis
LITERATURE REVIEW
The methods of three-dimensional analysis of slopes are usually extensions of conventional two-
dimensional approaches.
Variational calculus, for instance, has been extended to three-dimensional conditions by
Leshchinsky et al. (1985) and Leshchinsky and Baker (1986). Leshchinsky and Huang (1992)
further extended their original work, but the method was limited to problems with symmetric
geometry. Michalowski (1989) presented a three-dimensional solution based on the upperbound
theorem. The solution was limited to homogeneous slopes.
More recently, Farzaneh and Askari (2003) have extended the work by Michalowski (1989) to non
homogeneous slopes. Chen et al (2001a, 2001b) have also presented an upper-bound solution for
three-dimensional slope stability. Lam and Fredlund (1993) have presented an extension of the
GLE limit equilibrium method to three-dimensional conditions.
Other modeling approaches have been presented by numerous researchers in the last few years.
The upper and lower-bound theorems have been applied along with the finite element method, in
order to produce stress and strain fields (Lyamin and Sloan, 2002a and 2002b).
From the point of view of practicing geotechnical engineers, it becomes difficult to determine what
three-dimensional method of slope stability analysis is the more adequate. A sound theoretical
basis, a generalized approach that is capable of handling field conditions, and simplicity, are some
of the requirements of a handy slope stability method. It appears that if a practical three-
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis
THEORY
The state of stress and pore-water pressure at any point in the soil volume may be determined
using the finite element method. The computation of the factor of safety can be summarized as
follows:
a) The distribution of stresses and pore-water pressures are determined using the finite element
method. Appropriate boundary conditions, constitutive models, and constitutive parameters
must be adopted;
b) The normal and shear stresses are computed for a grid of points located at the base of the slip
surface. The normal stress depends on the position along the slip surface. The shear stress
depends not only on the position at the slip surface but also on the direction of slippage
projected on the horizontal plane;
c) Integration of the acting and resisting stresses is performed along the slip surface area
) Spherical and ellipsoidal slip surface shapes have been implemented
) Given the computed n, the shear strength can be calculated using the MohrCoulomb criterion for
saturated/unsaturated soils
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis

Finite Element models usually employ procedures based

on stresses that are computed at the integration points
The state of stress determined at the integration points
must be used.
If necessary, these stresses can be extrapolated to the
nodes using simple mapping techniques.
Several optimization techniques are available for the
determination of the critical slip surface. This paper will
not deal with these procedures.
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis
There are few numerical modeling tools available. Most of
the available tools are not practical and efficient enough
for routine use.
Two benchmark problems have been selected for the
verification of the threedimensional slope stability
analysis solution. The first problem corresponds to a
simple and symmetric cohesive slope. The second
problem corresponds to an asymmetric slope with friction
and cohesion. Both problems have been frequently
presented in the research literature for benchmark
purposes.
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis
Symmetric Cohesive Slope
A spherical slip surface is employed. The simple geometry, boundary conditions and
soil properties allowed for the development of analytical solutions. Baligh and Azzouz
(1975) and Gens et al. (1988) present two different solutions. Hungr et al. (1989),
Lam and Fredlund (1993) and Chen et al. (2001) have also analyzed this problem.
The parameters adopted herein for the stress analysis where as follows: a Young
Modulus ; Poissons ratio that varied from 0.1 to 0.49; total cohesion of 0.1 kPa,
friction angle equal to zero; pore-water pressure equal to zero; and unit weight of 1
kN/m3.
The simple geometry and absence of external loads results in smooth contours for
the distribution of stresses. However, the results depend on the value of Poissons
ratio. Values of factor of safety as high as 1.438 were obtained when increasing
Poissons ratio near its maximum theoretical value of 0.5.
A variation of 0.05 in the factor of safety was obtained when subjecting the analysis
to extreme variations in the number of nodes.
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis

Non-symmetrical slope with friction and cohesion

Leshchinsky et al. (1985) have proposed an analytical solution for three-
dimensional slope stability problems using the logarithmic spiral.
The same problem was also analyzed by Hungr et al. (1989) and Stianson
(2006), using different approaches.
Once again, the simple geometry and absence of external loads results in
smooth contours for the distribution of stresses. The differences in factor of
safety among the three previous researchers are due to differences in the
method of analysis and, more importantly, differences in the position of the
slip surfaces obtained.
Poissons ratio appears to have an effect on the factor of safety. Higher
Poissons ratios result in higher factors of safety.
4.Three-Dimensional Slope Stability Model
Using Finite Element Stress Analysis

A practical three-dimensional slope stability approach

was presented, using simple finite element stress and
seepage analyses.
Higher values of Poissons ratio resulted in higher values
of factor of safety
5.Numerical Modelling of Soil-Atmosphere
Interaction for Unsaturated Surfaces
The soil-atmosphere interaction processes can be represented by a series
of partial differential equations. This paper presents a PDE formulation that
was developed for soil-atmosphere analysis and presents three cases
demonstrating the application of the formulation developed to laboratory
and fields conditions.
Comparisons against experimental data show that evaporative fluxes can
be successfully reproduced by theoretical models. The PDE solutions were
used for the simulation of the fluxes through two soil cover configurations
to exemplify the application of theoretical models to design.
The results indicate that the manner how runoff is computed strongly
affects the results. The numerical solutions appear robust and can be
applied to the design of soil structures such as soil-cover systems, geo-
hazard hazard quantification, and other unsaturated soil problems.
5.Numerical Modelling of Soil-Atmosphere
Interaction for Unsaturated Surfaces

THEORETICAL FRAMEWORK
METHODOLOGY
ANALYSIS OF EVAPORATION FROM A SOIL COLUMN
LABORATORY TEST
ANALYSIS OF TWO SOIL COVER CONFIGURATIONS
5.Numerical Modelling of Soil-Atmosphere
Interaction for Unsaturated Surfaces
This paper presented a PDE formulation that was developed for
soil-atmosphere analysis and presented three cases
demonstrating the application of the formulation developed to
laboratory and fields conditions.
Comparisons against experimental data showed that evaporative
fluxes can be successfully reproduced by theoretical models. The
PDE solutions were used for the simulation of the fluxes through
two soil cover configurations.
The results indicate that the manner how runoff is computed
strongly affects the results. The numerical solutions appear
robust and can be applied to the design of soil structures such as
soil-cover systems.
6.Analysis of water seepage through earthen
structures using particulate approach(lucrare de
master~doctorat)
ANALYSIS OF TRANSIENT WATER SEEPAGE IN A
PAVEMENT SYSTEM USING THE PARTICULATE APPROACH
TRANSIENT SEEPAGE MODEL FOR PARTLY SATURATED
AND SATURATED SOILS USING THE PARTICULATE
APPROACH
7.Application of dynamic programming to evaluate the
slope stability of a vertical extension to a balefill
8.Earthen Embankment Breaching
 9.The implementation of unsaturated
soil mechanics into geotechnical
engineering
10.Groundwater Modelling Software
Capabilities and Limitations