Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Arabian Journal of Geosciences (2022) 15: 152

https://doi.org/10.1007/s12517-022-09454-z

ORIGINAL PAPER

Active lateral pressure to rigid retaining walls in the presence


of an adjacent rock mass
Mahdi Maleki1 · Meysam Imani1

Received: 28 January 2021 / Accepted: 6 January 2022 / Published online: 12 January 2022
© Saudi Society for Geosciences 2022

Abstract
If there is not adequate space for development of failure wedge behind a retaining wall, the conventional methods of deter-
mining the active earth pressure cannot be used. In such cases, the wall design procedure changes due to the change in the
distribution of the soil stress in the active state and the variations of the magnitude and the position of the resultant force
acting on the wall. In the present study, finite element method was employed to analyze the lateral earth pressure behind
a wall, considering the limitation on the backfill width. A rock mass with an unlimited width was considered behind the
limited backfill. In the analyses, the effects of various strength parameters of the backfill behind the wall and the rock mass
behind the backfill were taken into account. Besides, different wall geometries and backfill dimensions were analyzed. The
obtained results show that the earth pressure due to the limited backfill is considerably smaller than the lateral pressure
resulting from Coulomb’s method with an infinite backfill width. Also, the position of the active resultant earth force is
higher than one-third of the wall height.

Keywords Lateral earth pressure · Rock mass · Hoek–Brown · Retaining wall · Limited backfill

Introduction distribution of active earth pressure behind retaining walls


with limited backfill width. Among others, Frydman and
Rankine and Coulomb methods are commonly used for Keissar (1987) conducted a series of centrifuge tests on
determining the lateral pressure applied to the walls. Both rigid retaining walls filled with granular material to observe
methods assume that the width of the backfill behind the changes in lateral pressure behind the retaining wall from
wall has no limitation and the failure surface can fully be at rest to active condition. According to the results, they
developed. If the width of backfill is limited, the full devel- found that the coefficient of the active earth pressure (Ka)
opment of the failure surface is impossible and these the- decreases with increasing the depth of the wall. They also
ories are not inapplicable (Tsagareli 1965; Yang and Yin concluded that the Ka for the case of a limited backfill is
2006). Some practical applications of the limited backfill lower than the Ka calculated by Rankine’s theory. The cen-
behind a wall were shown in Fig. 1. For estimating active trifuge tests performed Take and Valsangkar (2001) showed
lateral pressure in the case of a limited backfill, analytical that the lateral earth pressure in the case of limited backfill
studies have been conducted (Spangler and Handy 1984, is lower than the magnitude calculated using the Coulomb
Chen et al. 2016 and Greco 2013). In addition to the ana- and Rankine classical earth pressure theories. As a general
lytical method, there are few experiments to investigate the conclusion from centrifuge tests performed by different
researchers, one can conclude that the traditional methods,
like the Rankine method, are not appropriate for the case of
Responsible Editor: Murat Karakus
limited backfills. In such cases, new methods should be used
* Meysam Imani to anticipate the lateral earth pressure, correctly. Therefore,
Imani@aut.ac.ir numerical method was applied by researchers to deal with
Mahdi Maleki such condition using Finite Element Analysis, one of appli-
mahdi1793@aut.ac.ir cations of finite element analysis include the detection of
instability mechanism of limited backfills and the suggestion
1
Geotechnical Engineering Group, Amirkabir University of a reduction coefficients for active lateral pressure (Yang
of Technology, Garmsar Campus, Garmsar, Iran

13
Vol.:(0123456789)
152 Page 2 of 11 Arab J Geosci (2022) 15: 152

Fig. 1  Cases of application


of rigid retaining walls with
limited backfill width: (a) near
an existing road; (b) near a
rock face; (c) near an existing
building

and Liu 2007; Leshchinsky et al. 2004). Li et al. (2017) of the resultant active force will ascend with respect to the
used discrete element method to simulate the lateral force wall height. However, one of the limitations of this study
acting on rigid retaining walls in the case of limited back- was to consider the elastic behavior for the rock mass behind
fill. They found that the lateral force acting on the wall will the backfill. In the present study, it was tried to eliminate
be reduced when the horizontal displacement of the wall this limitation by considering the Hoek–Brown failure for
is reduced and eventually it will reach to a constant value. the rock mass behind the backfill. The effects of different
Chen et al. (2018) investigated the effect of the horizontal parameters such as the wall height, the internal friction angle
displacement of the wall on the active lateral pressure in of the backfill soil behind the wall, and various geomechani-
the case of a limited backfill. Their findings show that the cal parameters of the rock mass behind the backfill, on the
failure will occur only in the backfill soil, not through the distribution of the lateral pressure and the position of the
rock media. Leshchinsky et al. (2004), based on a series resultant force acting on the wall were also studied.
of limit equilibrium analyzes on segmental retaining walls
with a narrow width, found that increasing the ratio of wall
height to backfill width will reduce the coefficient of active Numerical modeling
lateral pressure. Fan and Fang (2010) used the finite ele-
ment method to study the value and position of the resultant The finite element modeling method
active earth pressure acting on a rigid retaining wall with
the limited backfill width. They found that the coefficient of In this paper, the finite element modeling method was
active lateral pressure (Ka) was significantly lower than the used to analyze the lateral earth pressure acting on a
Coulomb solution. In addition, they found that the position rigid retaining wall near the stable rock face. The overall

13
Arab J Geosci (2022) 15: 152 Page 3 of 11 152

geometry used in this paper is depicted in Fig. 2, in which, The equivalent Elastic and Mohr–Coulomb
bt represents the backfill width on top of the wall, that can parameters of the rock masses using the Hoek–
be obtained using Eq. (1) based on the backfill width at the Brown failure criterion
bottom of the wall (bb), the wall height (H) and the angle
between the rock face and the horizontal direction (β). Δ The latest version of the Hoek–Brown failure criterion is as
is also the horizontal movement of the wall required for follows:
mobilization of the active state. ( )a
𝜎
𝜎1 = 𝜎3 + 𝜎ci mb 3 + s (2)
(1)

bt = bb + H tan(90 − 𝛽 ) 𝜎ci

In which, σ1 and σ3 are the major and minor principal


The wall used in the analyses represents a rigid concrete stresses, respectively, σci is the uniaxial compressive strength
wall with a thickness of 60 cm and a penetration depth of of the intact rock and mb, s, and a are the Hoek–Brown param-
one meter in the bed soil. Also, it was assumed that no eters of the rock masses, which can be obtained as follows:
bending occurs during the movement of the wall. As the ( )
wall foundation does not have a considerable effect on
GSI - 100
mb = mi e 28 - 14D (3)
the active pressure behind the wall stem, the foundation
was ignored in the models and the stability of the walls ( )
(4)
GSI - 100
was ensured by considering an embedment depth. A same s=e 9 - 3D

assumption was made by Fan and Fang (2010). A linear


elastic model was used for the concrete material, while the e(−GSI∕15) − e(−20∕3)
backfill behind the wall was made of sand and follows the a = 0.5 + (5)
6
Mohr–Coulomb failure criterion. Five parameters, namely
E, ν, c, ϕ, and ψ, are required to define this failure crite- In which, mi is the Hoek–Brown constant for the intact rock,
rion. The dilation angle (ψ) of the non-cohesive backfill D is the rock mass disturbance factor, and GSI is the Geo-
material was considered to be equal to ψ = ϕ-30º, provide logical Strength Index of the rock mass. Hoek et al. (2002)
that ϕ > 30º (Bolton 1986). proposed the following relations for obtaining the equivalent
The Mohr–Coulomb failure criterion was also consid- c and ϕ parameters of the rock masses:
ered for the rock mass behind the backfill soil. However, [ ]( )a−1
since the Hoek–Brown failure criterion has more compli- 𝜎ci (1 + 2a)s + (1 − a)mb 𝜎3n s + mb 𝜎3n
c=
ance with the rock mass behavior, the Hoek–Brown param-
√ ( ( )a−1 )
(1 + a)(2 + a) 1 + 6amb s + mb 𝜎3n ∕((1 + a)(2 + a))
eters of the rock mass were converted to the equivalent
Mohr–Coulomb parameters using the procedure described (6)
in the following section. [ ( )a−1 ]
6amb s + mb 𝜎3n
𝜙 = sin−1 ( )a−1 (7)
2(1 + a)(2 + a) + 6amb s + mb 𝜎3n

where σ3n is obtained as follows:


𝜎3 max
𝜎3n =
𝜎ci (8)

In which, σ3max shows the maximum confining stress. Also,


the modulus of deformation and the Poisson's ratio of the rock
masses can be obtained using the following relations that
were proposed by Hoek et al. (2002) and Vasarhelyi (2009)
respectively.
)√ 𝜎
D
(
(9)
ci
Em (GPa) = 1 − .10((GSI−10)∕40)
2 100

𝜈m = −0.002GSI − 0.003mi + 0.457 (10)

Fig. 2  The overall geometry of the numerical models developed in


the present research

13
152 Page 4 of 11 Arab J Geosci (2022) 15: 152

Boundary condition =0.0005H

As suggested by Lees (2016), the minimum distance between


the excavation face and the model boundaries should be four
times the excavation height. This recommendation was used
here as the distance between the sides and bottom bounda-
ries of the model to the concrete wall. The horizontal and
vertical movements of the bottom boundary were kept fixed,
while the side boundaries were kept fixed only in the hori-
zontal direction. Figure 3 shows a schematic of the finite ele-
ment model of the problem. For mobilizing the active state
in the backfill material, the wall should translate away from
the backfill soil. As suggested by Fang and Ishibashi (1986)
and Fan and Fang (2010), the displacement required for the
full mobilization of the active state is equal to Δ = 0.0005H
(see Fig. 4). Therefore, in this paper, depending on the wall
height, different horizontal displacements were applied to
the wall to simulate the active state behind the wall. Moreo- Fig. 4  The method of application of horizontal displacement to the
ver, the contact surface between two different materials was wall for the creation of the active state in the backfill behind the wall
considered using a shear strength reduction factor. For the
wall/backfill and backfill/rock mass interface, this reduction
factor was considered equal to 0.9 and 1, respectively (Poty- et al. (2016). In the centrifuge test conducted by Frydman
ondy 1961). and Keissar (1987), a vertical rock mass near a wall with
a height of 8.5 m and a bottom width (bb in Fig. 2) of one
meter was considered. The space between the wall and the
Verifications rock mass was filled with fine sand. At the time of placing
sand behind the wall, the wall lateral movements were pre-
Considering the methods which can consider vented. The sand unit weight and internal angle of friction
the limitation of the backfill width were equal to 15.8 kN/m3 and 36º, respectively. The friction
angle between the wall and the sand was equal to 25º. In
The results of this research were compared to three available another study, Fan and Fang (2010) dealt with this problem
methods to ensure the accuracy of the analyses. These meth- using the finite element method. They considered the linear
ods comprise the centrifuge test carried out by Frydman and elastic behavior for the rock mass behind the backfill. This
Keissar (1987), the finite element analysis by Fan and Fang assumption is not ideal for rock mass behavior. However,
(2010), and the analytical methods which were proposed in the present study, this problem was modeled considering
by Spangler and Handy (1984), Greco (2013) and Chen the plastic behavior for the rock mass and the correspond-
ing equivalent shear strength parameters were determined
using the Hoek–Brown properties of the rock mass. Assum-
4H
ing a total height of 8.5 m, similar to the centrifuge tests
performed by Frydman and Keissar (1987), the backfill soil
layers behind the wall were modeled in several layers with
H a thickness of 1 m for each layer to attain a height of 7 m.
The thickness of the last backfill soil layer was considered
equal to 1.5 m to fill the entire wall height. Immediately
after modeling each soil layer, a linear elastic model was
4H
considered for it. However, after modeling the next layer, the
Mohr–Coulomb model was replaced for the previous layer.
4H
Table 1 presents the parameters used in the verification mod-
els. Using Eq. (9), the considered Hoek–Brown parameters,
i.e., σci, GSI, and D, results in Em = 2100 MPa which is equal
to the magnitude that was used by Fan and Fang (2010). The
rest of the parameters of Table 1 were selected based on Fan
Fig. 3  A schematic of the finite element model of the problem and Fang (2010).

13
Arab J Geosci (2022) 15: 152 Page 5 of 11 152

As discussed previously, a horizontal displacement equal h/ z


to the Δ = 0.0005 × 8500 = 4.25 mm was applied to the wall 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6
0
to keep out the backfill soil and create the active state behind 0.5
the wall. After analyzing the finite element model, the hori- 1
zontal stress exerted from the backfill material to the wall, 1.5

i.e., the σh, was obtained through the depth of the wall and it 2
2.5
was compared to the other available methods as was shown 3 H=8.5m; =90°; =25°
in Fig. 5. The results of this comparison were shown in a 3.5
=15.8 kN/m3; =36°
z/bb
normalized form in which, γ is the unit weight of the backfill 4
Spangler MG and Handy (1984); bb=1m
sand, and z is the depth of any given point from the top of 4.5
Frydman & Keissar (Bottom cell) (1987)
5
the wall. Based on this figure, the σh/γz ratio resulting from 5.5 Frydman & Keissar (Top cell) (1987)
the present study which considers the plastic behavior for the 6 Fan & Fang (2010)
rock mass is larger on the upper two meters of the wall than 6.5 Greco (2013)
the results obtained by Fan and Fang (2010), considering the 7
Chen et al. (2016)
7.5
elastic behavior of the rock mass. However, the variations of 8
Present study

the lateral pressure and its value are approximately equal to


each other in both methods with an increase in depth. The Fig. 5  Comparing the active lateral earth pressures resulting from dif-
overall trend of the results obtained from the present study ferent methods
complies with the results of Spangler and Handy (1984),
Greco (2013) and Chen et al. (2016). However, the numeri-
cal method used in the present study resulted in a smaller the backfill behind it, and the stresses perpendicular to it
lateral pressure. Also, both the methods used in the present were considered as the backfill lateral pressure. For instance,
study and that proposed by Fan and Fang (2010) resulted in Fig. 6 shows the distribution of the active earth pressure for
relatively smaller lateral pressures than the centrifuge test the case of β = 80º, bb = 0.5 m and H = 5 m. Other required
(Frydman and Keissar 1987). parameters were selected from Table 1. As can be seen, the
distribution of the Coulomb and Rankine active earth pres-
Considering the methods which cannot consider sure is triangular in the soil depth, while the numerical anal-
the limitation of the backfill width ysis performed in this paper shows a non-triangular distribu-
tion of the active pressure which is lower than the Coulomb
A comparison was carried out between the results of the and Rankine values in most parts of the wall height. Hence,
present paper with the Coulomb and Rankine classical earth based on the numerical method applied in the present study,
pressure theories which cannot consider the limitation of the the position of the resultant active earth force is approxi-
backfill width. Following the application of the 0.0005H mately at a distance of 1.88 m from the bottom of the wall,
displacement to the wall for the creation of the active state, which is higher than the position of the resultant force in
a section was drawn at the junction between the wall and Coulomb and Rankine’s methods which is in one-third of

Table 1  The parameters used Parameter Rock mass Backfill Concrete wall

Uniaxial compressive strength of intact rock, σci (MPa) 33 - -


Geological strength index, GSI 40 - -
Coefficient of intact rock, mi 5 - -
Disturbance coefficient of intact rock, D 0.7 - -
Unit weight, γ (kN/m3) 24 15.5 24
Cohesion, c (kN/m2) - 0 -
Friction angle, ϕ - 31º -
Dilatancy angle, ψ 0° 1º -
Interface friction angle between the concrete wall and the backfill, δ1 - 28º -
Interface friction angle between the rock face and the backfill, δ2 - 31º -
Modulus of elasticity, E (MPa) 2100 60 2100
Poisson’s ratio, ν 0.36 0.3 0.15
Flexural rigidity, EI (kN/m2 /m) - - 4.2 × ­105
Normal stiffness, EA (kN/m) - - 1.26 × ­107

13
152 Page 6 of 11 Arab J Geosci (2022) 15: 152

Horizontal pressure, (kN/m 2) Effect of the angle between the rock face β
0 5 10 15 20 25 30 35 40 and the horizontal direction
0
H=5m; bb=0.5m; =80°
0.1 =15.5 kN/m 3 =31° Figure 7 shows the effect of β on the ratio of the Ka obtained
0.2 The Rankine method from the present study, i.e., Eq. (11), to that obtained from
0.3
The Coulomb method
the Coulomb’s method. It can be seen that the present paper
Present study
results in larger Ka than the method presented by Fan and
D e pth , Z /H

0.4
Fang (2010), which assumed the rock mass to behave as an
0.5
elastic material. Therefore, it can be concluded that the rock
0.6
mass plastic behavior and the Hoek–Brown failure criterion
0.7 result in a larger Ka and also a larger lateral pressure than
0.8 the case of an elastic rock mass. For β > 60º, a considerable
0.9 decrease in the Ka/Ka(Coulomb) ratio was seen. The β = 60º
1
angle is near to angle of the failure surface defined by Cou-
lomb (45 + ϕ/2 = 60.5º). Moreover, the decrease in the dis-
Fig. 6  The distribution of the horizontal earth pressure versus depth tance between the rock face and the wall (bb) results in a
after the application of various horizontal displacements to the wall decrease in the Ka/Ka(Coulomb) ratio.
Figure 8 shows the variation of the h/H with respect to
the β, in which, h denotes the distance between the point
the height from the bottom of the wall (equal to 1.65 m from of application of the resultant active force and the bottom
the wall bottom). of the wall. It is clear that by increasing the β, the space
Having the resultant active force, one can easily obtain required for developing the triangular wedge behind the wall
the coefficient of the active pressure as follows: will reduce. Therefore, the stress distribution exerted to the
wall will no longer have a triangular form which results in
Pa
Ka = (11)
0.5𝛾H 2 (cos 𝛿)

where Pa is the resultant active force, γ is the unit weight of Table 2  The different geometrical parameters for the sensitivity anal-
ysis
the backfill material, H is the retaining wall height, and δ is
the angle of friction between the backfill soil and the wall, Parameter Value
which is equal to 28º according to Table 1. The Pa value Wall height, H (m) 3, 5, 8
resulting from the numerical model in this study is equal Angle between the rock mass and the horizon, βº 60, 70, 80, 90
to 171 kN. Therefore, using Eq. (11), the corresponding Ka Bottom distance between the rock mass and the wall, 0, 0.5, 1
is 0.242. However, using the relations that existed in the bb (m)
foundation engineering textbooks, the values of Ka from the
Coulomb and Rankine methods are equal to 0.286 and 0.32,
respectively. Therefore, the existence of the rock mass near 1.3
the wall results in decreasing the Ka. 1.2
1.1
1
0.9
Parametric study for geometrical 0.8
H=5m
Ka/Ka(Coulomb)

=15.5 kN/m3; =31°


and mechanical aspects 0.7 Ka(Coulomb)=0.286
0.6 Present study; bb=0 m
Parameters used 0.5 Present study; bb=0.5 m
0.4 Present study; bb=1 m
Sensitivity analyses were carried out to study the lateral 0.3 Fan & Fang (2010); bb=0 m
earth pressure considering different backfill widths. To con- 0.2 Fan & Fang (2010); bb=0.5m
0.1
sider the effect of the problem geometry, including the wall Fan & Fang (2010); bb=1m
0
height (H), the angle between the rock mass with the hori- 30 40 50 60 70 80 90
zontal direction (β) and the distance between the rock mass Inclination of rock faces ( °)
and the wall (bb), different values were considered as pre-
sented in Table 2. The strength parameters of the backfill soil Fig. 7  Variation of Ka/Ka(Coulomb) versus the angle between the rock
and the rock mass behind the backfill were listed in Table 1. face and the horizontal direction (β)

13
Arab J Geosci (2022) 15: 152 Page 7 of 11 152

ascending the position of the resultant active force. The 1.8 Fan & Fang (2010)
results show that for β < 60º, the lateral pressure from the 1.7 Y=Ka/Ka(Coulomb); X=bt/H
1.6 Y = 1.4314x3 - 3.9747x2 + 3.4688x + 0.0579;R² = 0.9393
backfill soil has a triangular distribution and the position of 1.5
1.4
the resultant active load is approximately in one-third of the 1.3

K a /K a ( C o u l o m b )
wall height from the bottom of the wall. Also, by receding 1.2
1.1
the rock mass from the wall, the position of the resultant 1
active load will come down. 0.9
0.8
0.7 Y=Ka/Ka(Coulomb); X=bt/H; Present study
Effect of the friction angle of the backfill soil 0.6 =25°;Y = 0.20x3 - 2.00x2 + 2.92x + 0.28; R² = 0.94
0.5
=31°; Y = 1.1975x3 - 3.387x2 + 3.11x + 0.2; R² = 0.93
0.4
Figure 9 depicts the effect of the friction angle of the backfill 0.3 =35°; Y = -0.81x3 - 0.44x2 + 1.77x + 0.23; R² = 0.89
0.2
soil on the Ka considering different values of bt/H, in which 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
bt is the upper width of the backfill soil. The curves shown bt/H
in this figure were drawn as the best lines passing through
the points resulting from several numerical analyses for any Fig. 9  Variation of Ka/Ka(Coulomb) versus bt/H for different values of
given friction angle for the backfill soil. Three values were the friction angle of the backfill soil
selected for the friction angles, which include 25°, 31°, and
35°. It is interesting to see that considering the rock mass by the rock mass behavior, i.e., either elastic or plastic. The
as an elastic material resulted in an equal Ka for all backfill figure reveals that for all considered ϕ and bt values, the
friction angles, as analyzed by Fan and Fang (2010), while resultant active load exerts in a range equal to 0.433 to 0.341
the curves obtained from the present study show that by times the wall height from the bottom of the wall.
considering the Hoek–Brown failure criterion for the rock
mass, the effect of the backfill soil friction angle on the Ka
can be sensible. Backfill soils with lower friction angle have Effect of the Hoek–Brown parameters
larger Ka, which means higher lateral pressure exerted to of the rock mass on the active pressure
the wall. Moreover, by considering the Hoek–Brown failure and the position of the resultant active force
criterion for the rock mass, the Ka and the corresponding
lateral pressure applied to the wall will increase with respect In order to study the effect of variation of the rock mass
to the assumption of the elastic behavior for the rock mass. properties on the active pressure, a retaining wall with a
This conclusion shows the importance of the results of the height of 5 m was considered. A backfill bottom width
current study in practical engineering applications. equal to 5 m and a rock mass with an angle equal to 80º
According to Fig. 10, the dispersion of the points showing with respect to the horizontal direction were assumed. The
the various friction angles of the backfill soil is not signifi- backfill and the retaining wall parameters were listed in
cant. Hence, it could be stated that the position of applica- Table 1, while the Hoek–Brown criterion parameters for
tion of the resultant active force is not significantly affected the rock mass were presented in Table 3. According to

1.1 Fan & Fang (2010)


0.6
H=5m 1 Y=h/H; X=bt/H
Y = 0.0876x3 + 0.1147x2 - 0.304x + 0.4553;R² = 0.7977
0.57 =15.5 kN/m3; =31º 0.9 Present study
0.54 Present study; bb=0 m 0.8 Y=h/H; X=bt/H
Y = -0.1298x3 + 0.3745x2 - 0.3901x + 0.4568; R² = 0.7307
0.51 Present study; bb=0.5 m 0.7 =25°
=31°
0.48 Present study; bb=1m 0.6 =35°
h/H

Fan & Fang (2010); bb=0 m 0.5


h/H

0.45
Fan & Fang (2010); bb=0.5 m 0.4
0.42
Fan & Fang (2010); bb=1m 0.3
0.39
0.2
0.36
0.1
0.33 0
0.3 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
30 40 50 60 70 80 90 bt/H
Inclination of rock faces ( º)
Fig. 10  Variations of the h/H versus bt/H for different values of the
Fig. 8  Variations of h/H versus the angle of rock face (β) friction angle of the backfill soil

13
152 Page 8 of 11 Arab J Geosci (2022) 15: 152

Table 3  The Hoek–Brown Parameter Value


0.9 Present study
parameters considered for the Y=Ka/Ka(Coulomb); X=GSI
rock mass GSI 10, 30, 50, 70, 90 0.89
Y = 1E-07x3 - 0.76E-05x2 - 5E-05x + 0.8676; R² = 0.8951
ci =5 MPa & D=0
σci (MPa) 5, 25, 50 ci =25 MPa & D=0
0.88 =50 MPa & D=0
mi 5, 10 ci

K a /K a ( C oul om b )
ci =5 MPa & D=0.7
D 0, 0.7 ci =25 MPa & D=0.7
0.87 ci =50 MPa & D=0.7

0.86

0.9 Present study


Y=Ka/Ka(Coulomb); X=GSI 0.85
Y = 1E-07x3 - 0.67E-05x2 - 8E-05x + 0.8696; R² = 0.8837
0.89
ci=5 MPa & D=0 0.84
ci=25 MPa & D=0 10 20 30 40 50 60 70 80 90
0.88 ci=50 MPa & D=0
GSI
K a /K a ( C oul om b )

ci=5 MPa & D=0.7


ci=25 MPa & D=0.7
0.87 ci=50 MPa & D=0.7
Fig. 12  Effect of GSI on the Ka/Ka(Coulomb) ratio, considering mi = 10
0.86

0.38
Present study
0.85
Y=h/H; X=GSI
Y = -8E-08x3 + 11.9E-06x2 - 0.0002x + 0.3703; R² = 0.8643 ci=5 MPa & D=0
ci=25 MPa & D=0
0.84 ci=50 MPa & D=0
ci=5 MPa & D=0.7
10 20 30 40 50 60 70 80 90 0.375
ci=25 MPa & D=0.7
GSI ci=50 MPa & D=0.7

Fig. 11  Effect of GSI on the Ka/Ka(Coulomb) ratio, considering mi = 5 0.37


h /H

0.365

the analyses in the previous sections, the active pressure


state behind the wall was created through the application
of displacement equal to 0.0005H to the wall. The other 0.36
10 20 30 40 50 60 70 80 90
hypotheses and the modeling method are similar to the GSI

previous sections.
Fig. 13  Effect of GSI on the h/H ratio, considering mi = 5
Effect of the geological strength index

Figures 11 and 12 show the effect of the geologi- Effect of the uniaxial compressive strength
cal strength index (GSI) on the K a /K a(Coulomb) ratio, of the intact rock (σci)
considering m i = 5 and 10, respectively. The trend of
variation of the Ka/Ka(Coulomb) in both figures is the same. The effect of σci on the Ka/Ka(Coulomb) ratio was shown in
In other words, as GSI increases from 10 to 50, the Figs. 15 and 16, while Figs. 17 and 18 indicate the effect of
K a/K a(Coulomb) decreases, while by increasing GSI from σci on the position of the resultant active force with respect to
50 to 90, this ratio increases. Therefore, the lowest mag- the wall height. Increasing the σci/γ × H results in decreasing
nitude of the active earth pressure occurs for the GSI = 50. the Ka/Ka(Coulomb) ratio and descending the position of the
The effect of GSI on the position of the resultant active resultant active load. Therefore, the corresponding active
force was also investigated. As can be seen from Figs. 13 pressure and the overturning moment applied to the wall
and 14, for both mi = 5 and 10, the highest level of appli- decrease.
cation of the resultant active force occurs at approxi-
mately 0.372 times the wall height which corresponds to
the GSI approximately equal to 50. Therefore, the most Effect of ­mi
critical condition for the wall overturning occurs in the
case of GSI = 50. It is interesting to note that the lowest The effect of mi on the Ka/Ka(Coulomb) and h/H was shown
location of the resultant force with respect to the wall in Figs. 19 and 20, respectively. As can be seen, the
height occurs in the case of the highest GSI. Ka/Ka(Coulomb) ratio decreases with an increase in mi, reducing

13
Arab J Geosci (2022) 15: 152 Page 9 of 11 152

0.38 Present study 0.885


ci=5 MPa & D=0 D=0.7
Y=h/H; X=GSI
D=0
ci =25 MPa & D=0
Y = -5E-08x3 + 4.6E-06x2 - 8E-05x + 0.3707; R² = 0.8961 0.88 GSI
ci =50 Mpa & D=0
90
ci =5 MPa & D=0.7
0.375 90
ci =25 MPa & D=0.7 0.875

ci =50 MPa & D=0.7

Ka/Ka(Coulomb)
0.87

10
h /H

0.37 0.865

30
0.86 30

70
0.365 70
0.855
50
50
0.85
40 135 230 325 420
ci / ( ×H)
0.36
10 20 30 40 50 60 70 80 90
GSI
Fig. 16  Effect of Ka/Ka(Coulomb) ratio on the σci / γH ratio, considering
mi = 10
Fig. 14  Effect of GSI on the h/H ratio, considering mi = 10

0.375
0.885 GSI D=0.7
D=0
D=0.7 50
D=0 50
0.88 30
GSI
30
90 10
0.875 90 0.37
70
Ka/Ka(Coulomb)

h/H

70
0.87

10 90
0.865
0.365 90
30
0.86 30
70
70
50
0.855
50

0.85 0.36
40 135 230 325 420
40 135 230 325 420
ci/ ( ×H) ci / ( ×H)

Fig. 15  Effect of Ka/Ka(Coulomb) ratio on the σci / γH ratio, considering Fig. 17  Effect of h/H on the σci / γH ratio, considering mi = 5
mi = 5

the lateral pressure acting on the wall. Besides, the position


of the resultant active force also moves down the wall.
0.375
D=0.7
GSI
D=0
50
50

Effect of disturbance coefficient (D) 30


30
10

0.37
Figure 21 shows the effect of the disturbance factor (D) on 70
70
the Ka/Ka(Coulomb) ratio. The figure reveals that by increas-
h/H

ing the disturbance of the rock mass, the Ka/Ka(Coulomb) ratio 90

and the corresponding active earth pressure increase. Also, 90


0.365
according to Fig. 22, by increasing D, the position of the
resultant active load rises with respect to the wall bottom.

0.36
Conclusions 40 135 230 325 420

ci / ( ×H )

The findings from this research are as follows:


Fig. 18  Effect of h/H on the σci / γH ratio, considering mi = 10

13
152 Page 10 of 11 Arab J Geosci (2022) 15: 152

0.375 0.885
ci =5 MPa
ci
=5 MPa
ci =25 MPa =25 MPa
GSI =50 MPa ci
ci
GSI ci
=50 MPa
50 0.88
50 90
50
90
30
0.37 30
0.875 90
30

Ka/Ka(Coulomb)
10
0.87 10
h/H

10
10 10
70 30
70 0.865 30
0.365 70 30
70
90
70
90 0.86
90
70
50
50
0.855
50
0.36
5 mi 10
0.85
0 D 0.7
Fig. 19  Effect of mi on the Ka/Ka(Coulomb) ratio, considering D = 0
Fig. 21  Effect of D on the Ka/Ka(Coulomb) ratio, considering mi = 5

0.375
=5 MPa 0.375
ci
ci =5 MPa
=25 MPa
ci
ci =25 MPa
GSI ci =50 MPa
GSI ci =50 MPa
50
50
50
50
50 50
30 30
30
0.37 30 0.37 30
30
30 30
30
10
10
10
h/H

10
h/H

10 70
70
70
70 70
70
70
0.365 90
0.365 70
90
90 90
90
90

0.36
0.36 0 0.7
mi D
5 10

Fig. 20  Effect of mi on the h/H ratio, considering D = 0 Fig. 22  Effect of D on the h/H ratio, considering mi = 5

1. Using Coulomb’s earth pressure theory results in over- wall, the resultant active load will be at the highest pos-
estimating the active earth pressure for the case of the sible position.
existence of a rock mass in vicinity of a rigid retaining 5. Increasing the friction angle of the backfill soil results
wall. in decreasing the coefficient of active earth pressure,
2. Considering the plastic behavior for the rock mass while it does not considerably change the position of the
behind the backfill soil results in increasing the coef- resultant load acting on the wall.
ficient of active earth pressure and ascending the posi- 6. As GSI increases from 10 to 50, the active earth pressure
tion of the resultant active load exerted to the wall with decreases and the position of the resultant active load
respect to considering the rock mass as an elastic mate- rise with respect to the bottom of the wall. However,
rial. for GSI > 50, the position of the resultant active load
3. By increasing the ratio of the backfill upper width to descends while the magnitude of the active earth pres-
its depth (bt/H), the coefficient of active earth pressure sure increases.
increases. 7. By increasing σci and mi, the coefficient of active earth
4. The position of the resultant active load concerning the pressure decreases, while the position of the resultant
wall bottom depends on the distance between the rock active load descends with respect to the bottom of the
mass and the wall. When the rock mass approaches the wall.

13
Arab J Geosci (2022) 15: 152 Page 11 of 11 152

8. Increasing the disturbance factor of the rock mass results Greco V (2013) Active thrust on retaining walls of narrow backfill
in increasing the coefficient of active earth pressure and width. Comput Geotech 50:66–78. https://​doi.​org/​10.​1016/j.​
compg​eo.​2012.​12.​007
also ascending the position of the resultant active load Hoek E, Carranza-Torres C, Corkum B (2002) Hoek-Brown failure
with respect to the bottom of the wall. criterion. NARMS-TAC, Toronto, pp 267–273
Lees A (2016) Geotechnical finite element analysis. ICE Publishing,
Westminster, pp 15–16
Leshchinsky D, Hu Y, Han J (2004) Limited reinforced space in seg-
Declarations mental retaining walls. Geotext Geomembr 22:543–553. https://​
doi.​org/​10.​1016/j.​geote​xmem.​2004.​04.​002
Conflict of interest The authors declare that they have no competing Li MG, Chen JJ, Wang JH (2017) Arching effect on lateral pres-
interests. sure of confined granular material: numerical and theoreti-
cal analysis. Granular Matter 19:20. https://​doi.​org/​10.​1007/​
s10035-​017-​0700-2
Potyondy JG (1961) Skin friction between various soils and construc-
References tion materials. Geotechnique 11:339–353. https://d​ oi.o​ rg/1​ 0.1​ 680/​
geot.​1961.​11.4.​339
Bolton MD (1986) The strength and dilatancy of sands. Geotechnique Spangler MG, Handy RL (1984) Soil engineering. Harper and Row,
36:65–78. https://​doi.​org/​10.​1680/​geot.​1986.​36.1.​65 NewYork
Chen FQ, Lin YJ, Li DY (2018) Solution to active earth pressure of Take WA, Valsangkar AJ (2001) Earth pressures on un- yielding retain-
narrow cohesionless backfill against rigid retaining walls under ing walls of narrow backfill width. Can Geotech 38:1220–1230.
translation mode. Soils Found 59:151–161. https://​doi.​org/​10.​ https://​doi.​org/​10.​1139/​t01-​063
1016/j.​sandf.​2018.​09.​010 Tsagareli ZV (1965) Experimental investigation of the pressure of a
Chen AM, Li M, Wang J (2016) Active earth pressure against rigid loose medium on retaining walls with a vertical back face and
retaining walls subjected to confined cohesionless soil. Int J horizontal backfill surface. Soil Mech Found Eng 2:197–200.
Geomech 17(6). https://​doi.​org/​10.​1061/​(ASCE)​GM.​1943-​5622.​ https://​doi.​org/​10.​1007/​bf017​06095
00008​55. Vasarhelyi B (2009) A possible method for estimating the poisson’s
Fan CC, Fang YS (2010) Numerical solution of active earth pressures rate values of the rock masses. Acta Geodaetica Geophys 44:313–
on rigid retaining walls built near rock faces. Comput Geotech 322. https://​doi.​org/​10.​1556/​AGeod.​44.​2009.3.4
37:1023–1029. https://​doi.​org/​10.​1016/j.​compg​eo.​2010.​08.​004 Yang KH, Liu CN (2007) Finite-element analysis of earth pressures for
Fang YS, Ishibashi I (1986) Static earth pressure pressures with various narrow retaining walls. Geo-Engineering 2:43–52. https://d​ oi.o​ rg/​
wall movements. ASCE 112:292–333. https://​doi.​org/​10.​1061/​ 10.​6310/​jog.​2007.​2(2).1
(ASCE)​0733-​9410(1986)​112:​3(317) Yang XL, Yin JH (2006) Estimation of seismic passive earth pressures
Frydman S, Keissar I (1987) Earth pressure on retaining walls near with nonlinear failure criterion. Eng Struct 28:342–348. https://​
rock faces. ASCE 113:586–599. https://​doi.​org/​10.​1061/​(ASCE)​ doi.​org/​10.​1016/j.​engst​ruct.​2005.​08.​007
0733-​9410(1987)​113:​6(586)

13

You might also like