Dams with concrete asphaltic core are one of important options in designing
embankment and rock fill dams, especially in areas that suffer from lack of
fine grain materials with good quality and appropriate quantity for constructing
clay core dams. Advantages of this kind of dams are lack of crack and appropriate
impediment, not being sensitive to different climate conditions, sluicing during
construction, self-repairing capacity of core material, flexibility and deformability,
durability and resistance against continuous seepage, relative resistance against
earthquake and high security in war conditions. Several cases of this type of
dam have been constructed over the world. Lefebvre and Duncan
(1974) studied some of the factors leading to transverse cracking in low-embankment
dams using finite element analysis. The factors studied were the analysis procedure,
gravity turn-on or construction sequence, the magnitude and time of occurrence
of the settlement of the dam, the stress-strain characteristics of the dam material
and the shape of the abutment profile. Chugh (1983) conducted
one-dimensional wave propagation method for earthquake response analysis of
horizontally-layered sites of infinite lateral extent is adapted to account
for the finite cross-sectional dimensions of an embankment dam overlying a foundation
deposit which may be considered infinite in its lateral extent. A two-dimensional
dynamic finite element analysis was also performed for that case. The comparisons
of computed and observed responses support the modified use of the simple numerical
procedure. Zhao et al. (1993) carried out a systematic
investigation into the effect of both the type of impervious members and the
reservoir bottom sediment on the dynamic response of embankment dams using the
finite and infinite element coupled method. Tancev and Kokalanov
(1995) developed an incremental, nonlinear finite element procedure, suitable
for deformation, stress and stability analysis of embankment dams with waterproof
elements other than earth. The procedure was applied for analysis of hypothetical
rock-fill dams with asphaltic facing and an internal asphaltic core -vertical
and inclined. Studies were carried out to understand the prototype behavior
of these types of rock-fill dams. Both rock-fill and asphalt behavior was modeled
by using hyperbolic relations. Akkose el al. (2007)
studied stochastic seismic response of a rock-fill dam by finite element method.
The Keban dam constructed in Elazig, Turkey was chosen as a numerical example.
The interaction of the rock-fill dam with the reservoir was neglected, but not
the foundation rock. Tanaka (2007) studied the elasto-plastic
and viscoplastic constitutive relations with kinematic strain hardening-softening
model. A generalized return-mapping algorithm was applied to solution methods
of the problems. The dynamic relaxation method for static problems and the dynamic
analysis for earthquake responses were solved based on finite element methods.
Moayed and Ramzanpour (2008) studied the dynamic behavior
of a zoned core earth-fill dam which due to lack of suitable clay materials,
the dam was designed as zoned core that was composed of three vertical zones
including Central Lean (CL) clay core and two sides clayey Gravel Layers (GC).
Seismic behavior of dam was analyzed in two cases including homogenous clayey
core and zoned core by finite element method. The results showed that the displacements,
accelerations and spectral response in simple core are more than zoned core.
Tsompanakis et al. (2009) focused on the simulation
of the seismic response of a typical embankment using artificial neural networks.
The dynamic response of the embankment was evaluated utilizing the finite-element
method, where the nonlinear behavior of the geo-materials can be taken into
account by an equivalent-linear procedure.
In this study finite element analyses have been performed in order to investigate
the behavior of Gabric embankment dam with asphaltic core which is located at
Hormozgan, Iran with 41 m height and 13 m thickness and is placed over alluvial
bed, during construction and sluicing stages. Asphaltic core has been located
over alluvial foundation and under it cutoff wall for controlling seepage of
water in foundation been considered. Locating asphaltic core over cutoff wall
and also dam's shell on alluvial foundation could cause asymmetric seepage during
construction and sluicing stages that can affects behavior of dams body.
Foundation of dam consists of two layers; overlaying layer is silty sand and
located higher than rivers level. Underlying layer is a mixture of silty
sand and gravel that are under river's level. However, during construction,
higher layer in up-stream has been removed completely, while in down-stream
it was removed partially. Also, the effect of cores deformation modulus,
stiffness of coarse grain alluvial foundation materials and amount of remove
of fine grain alluvial on the results has been investigated too.
GEOMETRY OF MODEL
Due to geometrical characteristics of dams location, plane strain condition in maximum cross section of dam is modeled with acceptable approximation (Fig. 1). This section has maximum height of asphaltic core and alluvial thickness. Displacement of bed rocks in all analyses are considered zero.
Gabric dam placed on Gabric River at Hormozgan, Iran has height of 41 m and is placed on an alluvium layer with 13 m thickness. Dams construction started in 2003 and finished in 2009. Static finite element analysis by assuming linear behavior for core material and cutoff wall and also non-linear behavior for other material in plane strain condition has been conducted. In this method, locating of dam construction and sluicing reservoir has been considered step by step. Figure 2 shows the meshing plan of the model.
Because of limitation of strain values in asphaltic core, linear elastic behavior
with acceptable approximation could be assumed for cores materials (Adikari
et al., 1988) For estimation of behavioral parameters of cores
material, offered values of previous studies have been used (Hoeg,
|| Gabric dam section
|| Finite element model of the dam
Also, for investigating the effect of changes in deformable modulus, analysis
has been conducted for both 100 and 250 MPa deformation modulus. In both cases,
passions ratios were 0.49. Cutoff wall also modeled as linear elastic
which elastic parameters have been selected based on (ICOLD,
1985) offering from material characteristic of alluvial foundation. Behavior
of other material including rock fill and shells, transition zones in up-stream
and down-stream parts of core, fine and coarse grain alluvial foundation have
been estimated by non-linear hyperbolic model. For studying the effect of stiffness
of alluvial foundation material on stresses and strains of asphaltic core, three
different values for elastic modulus parameter of core in hyperbolic model have
been considered that were K = 200, 400 and 600.
Stresses distribution and load transfer in various parts of dams body, lateral strain values in asphaltic core and occurrence of shear and tensile failure in asphaltic core have been presented.
Stress distribution: Vertical stress distribution in the end of dam construction has been shown in Fig. 3. In vertical stress distribution, strong increase in transition from shell and transition layers toward asphaltic core are observed which is because of considerable difference between stiffness and deformation modulus of asphaltic core and sand and gravel shell that leads to non-equal settlement of them and loading transfer from shell to core. Vertical stress values in asphaltic core reach two times of overburden pressure that indicates considerable load transfer to asphaltic core.
Displacements distribution: Displacements distribution of dams body and foundation in the end of construction has been shown in Fig. 4 and 5. In Fig. 4, vertical displacement contours in the end of construction have been depicted. Maximum settlement occurs in the middle of height at down-stream shell and adjacent to transition zone that is equal to 0.28 m and decrease rapidly toward slopes. Asymmetric vertical settlement distribution in up-stream and down-stream shell is associated with asymmetric material characteristics and dams geometry.
Horizontal displacement distribution in the end of dams construction has been shown in Fig. 5. Maximum horizontal displacement occurred in down-stream shell and is equal to 0.18 m downward. Maximum horizontal displacement in up-stream shell is equal to 0.08 m. Overall, horizontal displacement of dam in the end of construction has relatively small values and indicates the stable situation at this condition. Horizontal displacements after sluicing have been depicted in Fig. 6. In this condition points' shift is toward downward and maximum horizontal displacement in down-stream shell is equal to 0.24 m. Maximum vertical displacement of asphaltic core in the end of construction was 0.22 m. Horizontal displacement of asphaltic core in this case is negligible, while after sluicing it reaches 0.14 m in dams crown.
Control of failure and lateral strain in core: Lateral strain values
in asphaltic core elements in various cases have been shown in Fig.
7a-d. As it could be seen, lateral strain values in core
reaches 1.2%, thus there are no problems due to change in permeability value.
Maximum of shear stress values in asphaltic core in various cases has been shown
in Fig. 8a-c. Possibility of shear or tensile
failure in asphaltic core after sluicing due to results of compacted asphaltic
core material is low. Hoeg (1993) obtained similar results
for shear or tensile failure in asphaltic core.
Effect of cores deformation modulus: As mentioned before, for
estimation of behavioral parameters of cores materials, offered values
from previous researchers in similar projects have been used. Thus, analyses
have been conducted by two deformation values: 100 and 250 Mpa. Obtained results
include changes of stress and strain values and also displacements in core have
||Vertical stress contour in the end of construction
||Vertical displacement contour in the end of construction
||Horizontal displacement contour in the end of construction
||Horizontal displacement contour after sluicing
||(a-d) Variation of lateral strains in asphaltic corefig
||(a-c) Variation of stress in asphaltic core after sluicing
As it could be seen in Fig. 8, with increase in stiffness of core material, maximum principle stress in all point of core in depth has been increased that indicates more absorption of stress by asphaltic core. Also by increasing of stiffness of core material, maximum shear stress value in all points of core has been increased. In both cases, maximum shear stress values in asphaltic core show that the possibility of shear failure in core is very low. By studying vertical and horizontal strains in core, it can be concluded that by decreasing the stiffness of core material, vertical and horizontal strain would increase, although in all cases horizontal strain values reached 1.2% and as a result, asphaltic core after deformation remains impermeable.
Effect of fine grain alluvial removal in down-stream of dam: One of
important issues from economical point of view is the amount of alluvial removal.
For this purpose, several analyses for studying the effect of removal of fine
grain alluvia on results have been conducted. As it has been presented in Fig.
7 and 8, principally, total fine grain alluvial removal
from beneath of down-stream shell has no effect on the amount of strain and
stress in asphaltic core. Thus, removal of alluvial from beneath of down-stream
shell will not assist adjustment of stresses and strains in core.
Effect of alluvial foundations stiffness: The most important issue in stress-strain behavior of Gabric dam is considering the presence of alluvial foundation in stress and strain values in asphaltic core. Therefore, hyperbolic modulus parameter (K) for alluvial foundation within 200 - 600 m has been considered. In Fig. 8 stress variation in asphaltic core after sluicing in various states has been offered. As it could be seen, stress values in approximate depth of 20 m from dams crown (middle of dams height) in different states of alluvial foundation is approximately equal, but in depth deeper than 20 m (lower part of the dam) with decrease of stiffness of alluvial foundation, maximum principle stress values and shear stresses in asphaltic core elements has been increased. In Fig. 7, strain changes in asphaltic core after sluicing for different stiffness of alluvial foundation has been presented. As could be observed strain values in upper part of the dam in various states is approximately equal. But in lower part of the dam, with decrease in stiffness of alluvial foundation vertical and horizontal stresses have been increased.
Results of static stress-strain analyses for asphaltic core option of Gabric dam provide possibility of studying dams behavior in sluicing and construction stages. In these studies, the effect of core deformation modulus, stiffness of coarse grain alluvial foundation and removal of fine grain alluvial on principle stresses, principle shear stresses and vertical and horizontal strains have been investigated. Also, vertical stress distribution at the end of construction, vertical and horizontal displacement and horizontal displacements after sluicing have been presented. The following specific conclusions can be drawn from the study:
In various states, horizontal strain values in core elements are less than 1.2%, thus it could be anticipated that after deformation, asphaltic core will remain impermeable.
Regarding obtained shear stresses of core, possibility of shear failure in core elements is low.
By increasing the stiffness of core material, maximum principle stress in all point of core has been increased by increasing the depth that is an indication of more absorption of stress by asphaltic core.
By increase of stiffness of core material, maximum principle shear stress in all point of core has been increased by depth.
By decrease of stiffness, both vertical and horizontal strains have been increased.
Total removal of fine grain alluvial beneath of lower shell has negligible effect on stress and strain values in asphaltic core.
Values of stress and strain after sluicing in upper part of the dam in different stiffness of alluvial foundation were approximately equal, while in lower part with decrease of stiffness of alluvial foundation, maximum principle stresses and maximum shear stresses in asphaltic core elements have been increased.
Vertical and horizontal strain has been increased by decreasing of stiffness of alluvial foundation.