RESEARCH ARTICLE

Kun XIONG, Yunlong HE, Yunfeng PENG

Adaptability to geological faulted foundation of Hardfill dam
E
Higher Education Press and Springer-Verlag 2008
Abstract Hardfill dam is a new type of dam which has
the advantages of low stress level and even stress distri-
bution in a dam body, resulting in low demands to foun-
dations. Based on 2D linear elastic and elasto-plastic
calculations of gravity dam and Hardfill dam using finite
element method (FEM), the stress distribution in a dam
body and anti-sliding stabilization is analyzed on the geo-
logical faulted foundations with weak weathered rock and
soft interlayers. It is concluded that Hardfill dams have
better adaptability to geological faulted foundations than
gravity dams and is more secure and economically sound.
Keywords Hardfill dam, geological fault, dam founda-
tion, weak weathered rock, soft interlayer
1 Introduction
It has been proven in practice that there are always vary-
ing degrees of defects in natural foundations. With hydro-
power development, favorable geological conditions for
dams have been almost exhausted and the others that
remain usually have many problems, such as strong
weathered rock, soft zone, fault zone, etc. Therefore, the
geologically faulted foundations impact on the safety of
the dam has been given more attention by hydraulic engi-
neers. Usually, in order to satisfy the demands of the dam
on the foundation, some reinforcements have to be made.
When the flaws are on the surface, including weak weath-
ered rock, joints and soft interlayer, only a certain degree
of excavation can easily acquire a reliable and economical
effect; but when these defects are buried deep, it is clearly
not economical to cope with them by just relaying on
excavation. And measures such as consolidation grouting,
curtain grouting and anchor bolts are combined with it.
Moreover, in some cases, although the condition of
foundation is good enough to build a dam, conservative
design and engineering measures are still adopted due to
the uncertain security. Thus, if it is able to have an accur-
ate estimate on the stabilization of dam foundation, the
security and economic rationality can be obtained at the
same time.
The idea of Hardfill material and Hardfill dam was put
forward by Raphael [1] and Londe et al. [2]. The Hardfill
material is produced by adding water and a small quantity
of cement into river-bed sand and gravel or excavation
waste which can be obtained easily near the dam site.
And the Hardfill dam is constructed using Hardfill mater-
ial. Fromthe 1990s, the idea has been carried out widely in
Japan. The cemented sand & gravel (CSG) damming tech-
nology has been developed, applied and promoted [3].
As a new recommended type of dam, the prominent
advantage of the Hardfill dam is low stress level in dam
body, resulting in low demand on foundations [4,5]. On
the face of a geologically faulted foundation, if the
Hardfill dam can also have a high level of safety, little
to no reinforcing measures are needed and thereby cost
is reduced.
Meanwhile, with a shape intermediate between gravity
dam and concrete faced rockfill dam (CFRD), the struc-
ture of Hardfill dam is more like that of CFRD. The earth
and rockfill dam have good adaptability to various foun-
dations because of its local material for dam construction.
However, for little cement content in Hardfill material, it
is obvious that the modulus of elasticity of Hardfill mater-
ial is much smaller than common concrete, but still much
larger than rockfill. Thus, the adaptability to foundations
of Hardfill dam must not be as good as that of the earth
and rockfill dam. Compared with CFRD, the great
advantages of Hardfill dam are erosion resistance of the
dam material, the spillway is set on the dam body and the
low construction cost [6,7]. Therefore, only the gravity
dam is chosen to compare with the Hardfill dam in this
paper. Based on 2D linear elastic and elasto-plastic calcu-
lations of gravity dam and Hardfill dam with FEM, the
stress distribution in dam body and anti-sliding stabiliza-
tion are analyzed on the geologically faulted foundations
with weak weathered rock and soft interlayers.
Consequently, the adaptability to geologically faulted
foundation of Hardfill dam is studied.
Received December 14, 2007; accepted September 9, 2008
Kun XIONG, Yunlong HE (*), Yunfeng PENG
State Key Laboratory of Water Resources and Hydropower
Engineering Science, Wuhan University, Wuhan 430072, China
E-mail: [email protected]
Front. Archit. Civ. Eng. China 2008, 2(4): 343349
DOI 10.1007/s11709-008-0057-z
2 Characteristics of Hardfill material and
Hardfill dam
2.1 Characteristics of Hardfill dam
The typical profile shown in Fig. 1 reflects the character-
istics of Hardfill dam, which is symmetrical trapezoid-
shaped or approximately symmetrical. The shape of the
dam body is intermediate between gravity dam and
CFRD, and the dam slope can be determined from some
facts of the specific project, such as foundation condi-
tions, height of dam, performance of the filling material,
etc. The range of slope is commonly from 1:0.5 to 1:0.8.
The impervious concrete facing upstream acts as an
impervious barrier like CFRD. The spillway can be placed
on the dam body and water can flow over during the
construction period.
2.2 Mechanical properties of Hardfill material
The mechanical behavior will change along with the ag-
gregate gradation, the quantity of cement, and unitage of
water. In Ref. [8], the mechanical properties of Hardfill
material was tested. The main aggregate is river-bed sand
and gravel or excavation waste. Based on the results, the
typical stress-strain relationship is shown in Fig. 2, which
indicates that the Hardfill is a type of elastic-plastic mater-
ial. In the design the Hardfill dam is considered as elastic
mass and the design compressive strength is taken accord-
ing to the elastic range strength, not the peak strength.
Table 1 lists the mechanical properties of three Hardfill
material specimens, of which the unit quantity of cement
is 60 kg/m
3
and aged for 91 days. Because the unitage of
water is changed, the values of the properties given in this
table have a scope. It is obvious that the modulus of elasti-
city of Hardfill material is much smaller than common
concrete because of little cement content. But the modulus
of elasticity is still much larger than rockfills. Like the
modulus of elasticity, other mechanical properties are also
intermediate between concrete and rockfills. And accord-
ing to Refs. [8,9], the strength and elasticity modulus were
improved with the increase in cement content.
3 Adaptability to foundation with weak
weathered rock
Weak weathered rock is the most common defect in dam
foundations. Weathering induces transformation of the
rocks physical properties and the strength of the rock is
greatly reduced. In the past, due to the hope of placing the
dam on considerably fresh and integrated bedrock, the
entire weathered rock would be removed. This will not
only increase the amount of excavation and backfilling
concrete to increase the investment and prolong the con-
struction period, but also enlarge the area of water pres-
sure. Whats more, excessive excavation will bring the
fresh rocks stress relaxation owing to unloading. And
weathering damage occurs. Sometimes digging too deep
will worsen the stress state of the dam because of the large
stiffness of bedrock [10].
It is pointed out in Design Specification for Concrete
Gravity Dams (SL3192005) that, in principle, after cer-
tain reinforcements, the amount of excavation should be
reduced if the requirements for the dams strength and
stability are met. Through a great amount of engineering
practice and theoretical research, some experiences have
been gained in the use of weathered rock in dam founda-
tion [11]. For example, the lower weathered rock has been
partly used in Three Gorges Dam; in Er Tan project, the
lower weathered rock has been directly used, the mid
weathered rock has been used entirely after proper dis-
posal and the upper weathered rock has been partly used
after proper disposal.
3.1 Comparison of different dam base levels
Because of the trapezoid shape, the weight of Hardfill dam
increases by about 80% of that of gravity dam while the
area of the base increases by about 80%. As the upstream
slope results in more weight of water, the anti-sliding res-
istance greatly increases. Hence, the amount of excavation
Fig. 1 Typical profile of Hardfill dam
Fig. 2 Typical stress-strain curve of Hardfill material
344 Kun XIONG, et al.
will be reduced effectively to build a Hardfill dam rather
than a gravity dam on the foundation with weak weath-
ered rock of certain depth to achieve the economic and
security purposes.
Set up a favorable foundation without important faults
and fissures, but with some weak weathered rocks of cer-
tain depth which include the upper layer and the lower
layer of same thickness of 1 m. The material parameters
are listed in Table 2. The dam base level can be chosen on
the surface of upper weak weathered rock, the surface of
lower weak weathered rock or the surface of slightly
weathered rock, as shown in Fig. 3.
In the calculating with FEM, the height of the Hardfill
dam is 70 m and the crest width is 8 m, so is the gravity
dam. The calculation domain of the foundation extends
by 1.5 times the height of the dam in the upstream, down-
stream and the depth, and the boundaries of the founda-
tion are all constrained normally. Four-node isopara-
metric element is used.
Compared with gravity dam with a downstream slope
of 0.8, the upstream and downstream dam slope of
Hardfill dam are all 1:0.7. In analysis, the loads contain
the deadweight of the dam, water pressure on the
upstream face and the foundation uplift pressure. The
water level is 66 m upstream, and 0 downstream. It is
assumed that the uplift pressure to upstream water pres-
sure is 1/2 in the drain hole and water-tight diaphragm.
The uplift pressure is 0 at the toe of the dam and its
distribution is linear along the dam base. The material
parameters of the dam are also listed in Table 2.
Assuming seepage demands can be met at every base
level, the result of calculation shows that the demands of
deformation and compressed stress are also satisfied both
for gravity damand Hardfill dam. As no crucial faults and
fissures exist, the anti-sliding along the base level is the
main problem. According to the criterion, the shear for-
mula is chosen to calculate the safety factors. The mean
normal stress and shear stress of each element in the worst
sliding surface can be obtained with FEM first, and then
the anti-sliding safety factor is calculated with stress
algebraic sum method in term of Eq. (1)
K
0
~
(
P
n
i~1
s
i
Dl
i
)f
0
z(
P
n
i~1
Dl
i
)c
0
P
n
i~1
t
i
Dl
i
, 1
where s
i
, t
i
are the normal stress and shear stress of ele-
ment i on sliding surface respectively, Dl
i
is the length of
element i on sliding surface.
Table 3 shows the anti-sliding safety factors. According
to the criterion in Ref. [12], the safety factor should be
greater than or equal to 3.0 at the basic load combination,
regardless of the project grade. The result shows that, on
the same foundation, the anti-sliding safety factors of
70 m Hardfill dam is about 100% higher than gravity
dam with the same height. During the construction of
gravity dam, if the dam is built on the surface of weak
weathered rock, the safety factors are all less than 3.0,
which cannot guarantee its stability against sliding. And
the only choice is to build the dam on the surface of
slightly weathered rock, or use consolidation grouting to
reduce the excavation. For Hardfill dam, the stability
requirement can be met comfortably even if the dam is
built on the surface of upper weak weathered rock.
Consequently, the excavation is reduced effectively so as
to obtain economic benefits.
Table 1 Mechanical properties of Hardfill material
Hardfill material elasticity modulus/GPa compressive strength/MPa tensile stength/MPa
river-bed sand gravel 1 2.05.0 1.54.0 0.30.8
excavation waste 1.02.0 1.02.0 0.60.7
river-bed sand gravel 2 1.02.0 2.05.5 0.40.7
Table 2 Material properties for 2D linear elastic analysis
item elasticity
modulus/GPa
Poissons
ratio
unit mass/
kg?m
23
upper weak weathered rock 6 0.250 2000
lower weak weathered rock 8 0.250 2000
slightly weathered rock 10 0.250 2000
concrete 20 0.167 2400
Hardfill 2 0.200 2200
Fig. 3 Sketch map of choices of dam base level
Table 3 Anti-sliding safety factors according to three base levels
dam base level c9/MPa f9 K
gravity dam Hardfill
dam
surface of upper weak
weathered rock
0.3 0.7 1.98 4.44
surface of lower weak
weathered rock
0.5 0.8 2.65 5.87
surface of slightly weathered
rock
0.7 0.9 4.39 8.95
Adaptability to geological faulted foundation of Hardfill dam 345
3.2 Hardfill dam on weak weathered foundations with
different thickness and intensity
In the above, the stability of different dam base levels is
contrastingly analyzed for gravity dam and Hardfill dam
on weak weathered foundation with certain thickness. The
adaptability of Hardfill dam is obviously better than grav-
ity dam. But in fact the thickness and intensity of weak
weathered rock varies in the foundation. In the following,
using different thickness and strength of the weak weath-
ered rock foundation, the adaptability of Hardfill dam is
analyzed. The model of FEM is the same as above; how-
ever, the upper and lower weak weathered rocks are com-
bined to one layer with the thickness varying from 1 m to
20 m, the modulus of elasticity varies from 1 GPa to 8
GPa. Other parameters are the same as that in Table 2.
The result of FEM shows that, in each case of weak
weathered foundation with different thickness and intens-
ity, the stress distribution is quite even in the Hardfill dam.
The vertical normal stress is compressive stress and no
tensile stress occurs. Figures 4 and 5 show the curve of
s
1max
and s
3min
against the weathered rocks thickness
and intensity. If the calculation result is a positive value,
it means the stress is tensile, and vice versa.
In any case, s
1max
of the dam body always appears at
the dam heel. From Fig. 4, no matter what the thickness
is, s
1max
always increases along with the increase of the
rocks modulus of elasticity. During the case when the
thickness is 1 m and the elasticity modulus is 8 GPa,
s
1max
achieves its maximum value. Contrast to Table 1,
even in this case, the maximum still does not exceed the
design tensile strength of Hardfill material. In any case,
s
3min
of the dam body always appears in the middle of the
dam bottom. From Fig. 5, no matter what the thickness
is, s
3min
almost does not change. The minimum of s
3min
is
21.14 MPa. And contrast to Table 1, this value also does
not exceed the design compressive strength of Hardfill
material. Besides, the design compressive strength is taken
according to the elastic range strength, not the peak
strength, which hides some reserved strength.
Therefore, the trapezoid-shaped Hardfill dam has less
demands than gravity dam on weak weathered founda-
tion. In every case in this paper, the stress extremum of
dam body always remains in the range of design strength
of Hardfill material. The adaptability of Hardfill dam to
weak weathered foundation is favorable.
4 Adaptability to foundation with soft
interlayers
The unfavorable soft interlayer is also a common geo-
logical defect in foundations, which is usually the crucial
problem to the safety of the dam base. Therefore, in the
design of dams, it is not only needed to check the stability
against sliding along the base level, but also to verify the
possibility of the sliding along soft layers. Usually, there
are two types of the distribution, including single sliding
surface and double sliding surfaces. In this paper, the
foundation contains two soft interlayers.
4.1 Elasto-plastic model of foundation with soft interlayers
Set up a foundation containing two adverse soft inter-
layers, and the gravity dam and Hardfill dam of the same
height will be built on this base. In the 2D elasto-plastic
analysis with FEM, the dam structure, load conditions
and boundary conditions are the same as the example
above. The FEM mesh is shown in Fig. 6 and material
parameters are listed in Table 4.
The material properties are shown in Table 4. The con-
stitutive model is elasto-plastic and the yielding criterion
adopts Drucker-Prager criterion:
f ~aI
1
zJ
1=2
2
{H~0, 2
where I
1
is the first invariant, J
2
is the second stress devi-
ator invariant. And
a~
2 sin Q

3
p
3{sin Q
, 3
H~
6c cos Q

3
p
3{sin Q
, 4
where c is the cohesion,Q is the friction angle.
4.2 Comparison of different stress distributions
Figures 7 and 8 show the distribution of principal stress
s
1
, s
3
of gravity dam and Hardfill dam respectively. If the
Fig. 4 s
1max
against weathered rocks thickness and intensity
Fig. 5 s
3min
against weathered rocks thickness and intensity
346 Kun XIONG, et al.
Fig. 6 FEM mesh
(a) Gravity dam; (b) Hardfill dam
Table 4 Material properties for 2D elasto-plastic analysis
item modulus of elasticity/GPa Poissons ratio unit mass/kg?m
23
c/kPa Q/(u)
foundation 10 0.250 2000 700 42
concrete 20 0.167 2400 1 100 50
Hardfill 2 0.200 2200 500 45
soft interlayers 0.5 0.350 2000 20 30
Fig. 7 Distribution of principal stress s
1
/MPa
Fig. 8 Distribution of principal stress s
3
/MPa
Adaptability to geological faulted foundation of Hardfill dam 347
calculation result is a positive value, it means the stress is
tensile, vice versa.
From Fig. 7, the stress distribution of Hardfill dam is
more even. Especially in the position close to the base
level, the stress varied largely in gravity dam while the
stress almost does not change along the flow direction in
Hardfill dam, where the contour is nearly horizontal.
Shown in Fig. 8, the principal stress s
3
is also distributed
horizontally in Hardfill dam and does not show stress
concentrations which exist at the toe in gravity dam.
Table 5 shows the comparison of maximum stresses of
the dam body on the foundation with soft interlayers. The
maximum of principal stress s
1
in Hardfill dam is com-
pressive, which means that no tensile stress exists on the
whole profile. And the minimum of the principal stress s
3
is about 50% and the maximum of the horizontal shear
stress t in Hardfill dam is nearly 37% of that in gravity
dam. Obviously, the stress level has been reduced largely.
4.3 Comparison of anti-sliding safety
The overloading analysis is made to explore the integer
safety factors using the method of overloading the water
density [13]. And the integral safety of the dam is esti-
mated synthetically by the convergence of interactive pro-
cess and extending through of the plastic zone.
Figures 9(a) and 10(a) showthe distribution of the plas-
tic zone of gravity damand Hardfill damwith geologically
faulted foundation. The subscript 1 of K means before
reinforcement and 2 means after reinforcement. From
Fig. 9(a) of gravity dam, the plastic zone appears along
soft interlayers and almost links through when K is equal
to 1.0. As the overload factor is 1.9, the plastic zones along
the soft interlayers link through completely. On the same
foundation, there is also an area of plastic zone in the heel,
which is shown in Fig. 10(a). As the overload factor is 3.9,
the plastic zones just link through completely.
The overloading analysis is conducted again after
grouting and the soft interlayers are enhanced
(c 50.5 MPa, Q 540u). When the overload factor is 1.0,
there is no plastic zone in the gravity dam and the Hardfill
dam. Figures 9(b) and 10 (b) show the distribution of
plastic zone at the unstable period after reinforcement.
Due to the reinforcement, the integer safety factors have
been improved largely. When the dam is unstable, the
plastic zones extend through along the base of the gravity
dam and the Hardfill dam. There is also a great area of
plastic zone at the heel because of great water pressure.
The anti-sliding safety factors along the base level and
the soft interlayers are calculated separately according to
Eq. (1) and
K~
1
2
cH
0
H
0
1
2
c
0
H
0
H
0
~
c
c
0
~K
c
, 5
where K
c
is the over loading factor of the unit weight.
Table 6 shows the anti-sliding safety factors of the grav-
ity dam and the Hardfill dam with stress algebraic sum
method and overload method in the condition of natural
Table 5 Comparison of dam stresses/MPa
type of dam dam body
s
1
s
3
t
gravity dam 0.01 22.32 0.95
Hardfill dam 20.01 21.17 0.35
Fig. 9 Distribution of plastic zone of gravity dam
(a) K
1
51.0; (b) K
1
51.9; (c) K
2
53.3
Fig. 10 Distribution of plastic zone of Hardfill dam
(a) K
1
51.0; (b) K
1
53.9; (c) K
2
55.8
348 Kun XIONG, et al.
faulted foundation and reinforced foundation respect-
ively.
On the natural faulted foundation, the anti-sliding
safety factors along the base level are all above 3.0, sat-
isfying the demand of the criterion. But the anti-sliding
safety factor along the soft interlayers and the integer
safety factor of gravity dam are quite low. After some
reinforcements, the anti-sliding safety factor along the
soft interlayers is barely 3.12, which just meets the demand
of the criterion. To the Harfill dam before reinforcements,
the anti-sliding safety factor along the soft interlayers is
2.34 and the integer safety factor is already 3.9, which is
close to the usual control safety in the design. In fact, this
means that nearly no reinforcement is needed. Even if the
same reinforcements are applied, the safety factors of
Hardfill dam improve more. All of these reflect that
Hardfill dam has good adaptability to the foundation
with soft interlayers and simultaneously achieve security
and economic objectives.
5 Conclusions
Hardfill dam is a new type of dam which has some unique
advantages due to the even and low stress level. Based on
FEM, the adaptability to geologically faulted foundation
of Hardfill dam is studied. Some conclusions are drawn as
follows:
1) On the foundation with weak weathered rock of cer-
tain depth, it can reduce the amount of excavation effec-
tively and satisfy the demand of anti-sliding comfortably
to build a Hardfill dam rather than a gravity dam. The
purposes of economy and security can be achieved at the
same time.
2) In each case of weak weathered foundation with dif-
ferent thickness and intensity in certain scope, the stress
distribution is quite even in the Hardfill dam. The vertical
normal stress is compressive stress and no tensile stress
occurs. The stress extremum of dam body always remains
in the range of design strength of Hardfill material.
3) On the foundation with two soft interlayers, the dis-
tribution of stress is quite even and the stress level is rather
low in the body of the Hardfill dam. The anti-sliding
safety factors along the base level and soft interlayers
and the integer safety factor are about twice those of the
gravity dam. Compared with gravity dam, little to no
reinforcement is needed to build a Hardfill dam on this
type of foundation.
4) It is shown that the Hardfill dam has a better adapt-
ability to geologically faulted foundation with weak
weathered rock and soft interlayers of the types in this
paper.
Acknowledgements This study was supported by the National Natural
Science Foundation of China (Grant No. 50679058).
References
1. Raphael J M. The optimum gravity dam. In: Proceedings of
Roller Compacted Concrete III, ASCE, San Diego, California.
1992, 519
2. Londe P, Lino M. The faced symmetrical Hardfill dam: a new
concept for RCC. International Water Power and Dam
Construction, 1992, 44(2): 1924
3. Hirose T. Design concept of trapezoid-shaped CSG dam. In:
Proceedings of the 4th International Symposium on Roller
Compacted Concrete Dams, Madrid. 2003, 457464
4. Peng Yunfeng, He Yunlong, Xiong Kun. Study on the struc-
ture safety of Hardfill dam. China Water Resources, 2007,
(21): 5557 (in Chinese)
5. Peng Yunfeng, He Yunlong, Wan Biao. Hardfill dama new
design RCC dam. Water Power, 2008, 34(2): 6163 (in
Chinese)
6. Wang Xiaoqiang. Material Dissolution and Engineering
Application of Cement Sand and Gravel (CSG) Dam. China
Institute of Water Resources and Hydropower Research,
2005:45 (in Chinese)
7. Lu Shuyuan, Tang Xinjun. Brief introduction to a new type of
dam called face-plate-cementing rockfill dam. Journal of
Yangtze River Scientific Research Institute, 1998, 15(2): 54
56 (in Chinese)
8. Hirose T. Concept of CSG and its material properties. In:
Proceedings of the 4th International Symposium on Roller
Compacted Concrete Dams, Madri. 2003, 465473
9. Tang Xinjun, Lu Shuyuan. Preliminary research on mech-
anical behaviors of cemented rockfill material. Journal of
Wuhan University of Hydraulic and Electric Engineering,
1997, 30(6): 1518 (in Chinese)
10. Wang Tao, Wu Xubin, Zhu Jianmin. Study on selection of
foundation base of gravity dam. Hubei Water Power, 2004,
(3): 3032 (in Chinese)
11. Yu Bo, Xu Guangxiang. Thinking about research status &
development of dam foundation rock for hydropower pro-
jects. GuiZhou Water Power, 2007, 21(1): 17 (in Chinese)
12. Design Specification for Concrete Gravity Dams (SL-319-
2005), 2005 (in Chinese)
13. Xiong Kun, He Yunlong, Xiao Wei. Numerical analysis of
stability of Sanglang arch dam. Engineering Journal of
Wuhan University, 2007, 40(4): 2225 (in Chinese)
Table 6 Safety factors with three methods
before reinforcement after reinforcement
gravity dam Hardfill dam gravity dam Hardfill dam
along base level 3.12 7.81 3.10 7.60
along soft interlayer 1.47 2.34 3.12 6.16
integer safety factor 1.90 3.90 3.30 5.80
Adaptability to geological faulted foundation of Hardfill dam 349