Geomechanics and Engineering, Vol. 7, No.

3 (2014) 263-277
DOI: http://dx.doi.org/10.12989/gae.2014.7.3.263 263

Effect of seismic acceleration directions on dynamic earth
pressures in retaining structures
Ting-Kai Nian 1,2, Bo Liu 1, Jie Han 3 and Run-Qiu Huang 2
1
School of Civil Engineering & State Key Laboratory of Coastal and Offshore Engineering,
Dalian University of Technology, Dalian 116024, China
2
State Key Laboratory of Geohazard Prevention and Geoenvironmental Protection,
Chengdu University of Technology, Chengdu 610059, China
3
Department of Civil, Environmental and Architectural Engineering, the University of Kansas,
Lawrence, KS 66045, USA

(Received February 24, 2013, Revised April 22, 2014, Accepted May 16, 2014)

Abstract. In the conventional design of retaining structures in a seismic zone, seismic inertia forces are
commonly assumed to act upwards and towards the wall facing to cause a maximum active thrust or act
upwards and towards the backfill to cause a minimum passive resistance. However, under certain
circumstances this design approach might underestimate the dynamic active thrust or overestimate the
dynamic passive resistance acting on a rigid retaining structure. In this study, a new analytical method for
dynamic active and passive forces in c- soils with an infinite slope was proposed based on the Rankine
earth pressure theory and the Mohr-Coulomb yield criterion, to investigate the influence of seismic inertia
force directions on the total active and passive forces. Four combinations of seismic acceleration with both
vertical (upwards or downwards) and horizontal (towards the wall or backfill) directions, were considered. A
series of dimensionless dynamic active and passive force charts were developed to evaluate the key
influence factors, such as backfill inclination , dimensionless cohesion c/ H, friction angle , horizontal
and vertical seismic coefficients, kh and kv. A comparative study shows that a combination of downward and
towards-the-wall seismic inertia forces causes a maximum active thrust while a combination of upward and
towards-the-wall seismic inertia forces causes a minimum passive resistance. This finding is recommended
for use in the design of retaining structures in a seismic zone.

Keywords: earth pressure; retaining structures; analytical solution; horizontal and vertical seismic
coefficients

1.Introduction

The failures of retaining structures as earth retaining walls and stabilizing piles against sliding
under seismic loading have been reported by several researchers (Seed and Whitman 1970, Wood
1973, Fang and Chen 1995, Ling and Leshchinsky 1998, Das 2008). Recently, Yao et al. (2009)
investigated the failure of retaining structures adjacent to slopes after the magnitude 8.0 Wenchuan
Earthquake of China in 2008. They compiled different modes of earthquake-induced failures of
retaining structures, which include large inclination of stabilizing piles, dislocation, outward

Corresponding author, Associate Professor, E-mail: tknian@dlut.edu.cn

Copyright 2014 Techno-Press, Ltd.

http://www.techno-press.org/?journal=gae&subpage=7 ISSN: 2005-307X (Print), 2092-6219 (Online)
264 Ting-Kai Nian, Bo Liu, Jie Han and Run-Qiu Huang

deformation, overall failure of mortar stone facing retaining walls, mid-height shear failure or
horizontal sliding at the bottom of concrete retaining walls, embankment failure due to the
overturning of retaining walls, and global failure of reinforced earth retaining walls, etc. Sliding
and overturning of retaining structures are two common failure modes observed after an
earthquake, which result from the increase of lateral earth pressure. Clearly, dynamic earth
pressure is an important parameter for controlling the stability of earth retaining structures in a
seismic zone.
Currently, two earth pressure theories are widely adopted to calculate the dynamic earth
pressures in earth retaining structures, which are based on the Coulomb sliding wedge concept
(Mononobe 1924, Okabe 1924, Kapila 1962, Seed and Whitman 1970, Fang and Chen 1995,
Choudhury and Nimbalkar 2005, Ghosh 2008, Shukla et al. 2009, Shukla and Habibi 2011) and
the Rankine limit stress state (Richards et al. 1990, Budhu and Al-Karni 1993, Richards and Shi
1994, Lancellotta 2002 and 2007). The past studies have investigated cohesionless and c- backfill
(c is cohesion and is friction angle) with, horizontal and inclined ground surfaces, and under
horizontal and vertical seismic loading conditions. However, the direction effect of horizontal and
vertical seismic inertia forces and seismic earth pressures on retaining structures under a
combination of horizontal (towards the wall or backfill) and vertical (upward or downward)
seismic inertia forces have not been well studied (Fang and Chen 1995, Shukla and Habibi 2011).
In the conventional design, a combination of upward and towards-the-wall seismic inertial forces
is assumed to cause a maximum active thrust (Mononobe 1924, Okabe 1924, Fang and Chen
1995), while a combination of upward and towards-the-backfill seismic inertial forces is assumed
to cause a minimum passive resistance (Kapila 1962, Fang and Chen 1995, Shukla and Habibi
2011). Nian and Han (2013) investigated the seismic active earth pressure in c- soil with an
infinite slope. In fact, it is not clear under which combination of the seismic loading directions the
retaining structure is in the most dangerous state of instability and how to properly consider this
critical state in the design of retaining structures in a seismic zone. To the authors knowledge,
under certain circumstances the conventional design in which seismic inertia forces act upwards
and towards the wall (i.e., an active state), may underestimate the seismic active thrust while it
may overestimate the seismic passive resistance when the seismic inertia forces act upwards and
towards the backfill (i.e., a passive state). In the present study, the analytical solutions for seismic
lateral earth pressure/force were obtained based on the Rankine earth pressure theory, and then
they are used to investigate the influence of the seismic acceleration directions on active and
passive forces acting on earth retaining structures in a seismic active zone.

2. Analytical formulation

Fig. 1(a) shows a typical soil slice ABCD with a height z and a unit width 1 in an infinite slope
with an angle . The base of the soil slice is parallel to the slope surface. The effect of the left-side
soil mass of the cross section (i.e., AB) in the infinite slope can be replaced by a rigid retaining
structure such as a retaining wall or a row of stabilizing piles against sliding, which is rather
common in practice. In this study, the AB surface is assumed to be vertical and smooth. In a limit
equilibrium state under self-weight and seismic inertia forces in the semi-infinite mass of the c-
soil, the inclined lateral pressure value at the interface AB is the seismic active or passive earth
pressure acting on a retaining structure of cohesive backfill with an infinite top slope. Herein, the
 Effect of seismic acceleration directions on dynamic earth pressures in retaining structures 265

1
A D
c, ,
kvz z Pae(pe)
khz
ae( pe)

z
ae(pe) ae(pe)
v C
ae(pe) B
Retaining
Wall
ae(pe) v

(a)

z khz
xz
xz
x (1-kv)z

ae(pe) z
kvz

(b) (c)
Fig. 1 Force analysis of the soil slice in an infinite slopes under seismic loading: (a) Forces on
the soil slice; (b) Stresses on the wedge element; (c) Definition of the direction angle
induced by horizontal and vertical seismic inertia forces (Modified from Nian and Han
2013)

seismic active or passive earth pressure on the retaining wall is assumed to be parallel to the
inclined backfill surface.
Prior to the stress analysis, the following assumptions were made:
(1) the soil-wall interface is vertical and smooth;
(2) the backfill is cohesionless or c- soil;
(3) the ground surface is horizontal or inclined;
(4) the self-weight of the soil slice is z, where is the unit weight of the soil;
(5) the seismic inertia forces on the soil slice are horizontal (towards the wall facing or
backfill) and vertical (upwards or downwards), i.e., kh z [ + / ] and kv z [ + / ],
where kh and kv are the horizontal and vertical seismic coefficients respectively;
(6) the seismic direction angle of horizontal and vertical seismic inertia forces is defined as
kh
tan 1 , in which the positive sign + and the negative sign before the
1 k
v
term kv indicate that the seismic inertia forces act downwards and upwards, respectively
while the positive sign + and the negative sign before the term kh indicate that the
seismic inertia forces act towards the wall facing and the backfill, respectively (Figs. 1(a)
and (c)).
The normal stress and shear stresses obtained from the force equilibrium can be substituted
266 Ting-Kai Nian, Bo Liu, Jie Han and Run-Qiu Huang

into those from the wedge stress analysis to obtain the vertical stress z and shear stress xz which
satisfy the equilibrium and are expressed by a combination of the parameters: x, z, 1 kv, tan
and tan (x is an unknown variable). Furthermore, the vertical stress z and shear stress xz can
be substituted into the following relationship to obtain the major and minor principal stresses
(Nian and Han 2013)
1 1
2
1
z x z x xz2 (1)

3 2 2

These principal stresses (1, 3) are a function of x and satisfy both the equilibrium within the
soil domain of the slope and the stress boundary conditions. According to the lower-bound limit
analysis concept (Chen 2007), this stress field (1, 3) is statically allowable if it nowhere violates
the yield condition, such as the Mohr-Coulomb criterion. When the principal stresses (1, 3) are
substituted into the following Mohr-Coulomb failure criterion (Nian and Han 2013)

1 1
( 1 3 ) ( 1 3 ) sin c cos (2)
2 2

a quadratic equation as a function of x can be obtained, in which c and are cohesion and friction
angle of the soil, respectively. Solving this equation yields two principal values of the horizontal
stress x in a statically allowable stress field. The components x / cos are the seismic active and
passive earth pressures (ae and pe) on the retaining wall of c- backfill with an infinite top slope
as expressed below
4 tan 2
J 2 (1 A) 2
pe x cos 2


cos (1 k v )z cos J 4D 2D 2
(3a)
ae tan (1 A) ( )
1 kv 1 k v
or
pe x

cos (1 k v )z K (3b)
ae
where

K pe 4 tan 2 4D 2D 2
K cos J J 2 (1 A) 2
tan (1 A) ( ) (3c)
K ae cos 2 1 k v 1 kv

2 cos 2 1 A 2 D
J 1 A 1 tan (3d)
cos 2
1 A 1 kv

c
A tan tan , D , z0 (3e)
z

K is the seismic lateral pressure coefficient, Kae and Kpe are the seismic active and passive earth
 Effect of seismic acceleration directions on dynamic earth pressures in retaining structures 267

pressure coefficients, respectively, and the other parameters are defined earlier. The selection of
or + sign for the optional sign before the square root in Eq. (3) implies the active or
passive state of limit equilibrium. Especially, the analytical expressions of the seismic active and
passive earth pressures (ae and pe) at z = 0 can be rewritten as

pe x

cos 2c tan 2c tan 2 4c 2 cos
ae
(4)
sin 1
2c cos
cos

To check the reasonableness of the analytical solution, a special case that has a cohesionless
backfill with a horizontal ground surface under seismic loading is adopted, that is, = 0, c = 0,
0, kh 0, and kv 0. Eq. (3c) can be reduced to the following equation since = 0 and c = 0

K pe 1 sin 2 2
K
tan 2 tan 2 (5a)
K ae cos
2
cos

Eq. (5a) is the same as that presented by several investigators (Richards et al. 1990, Budhu and
Al-Karni 1993, Richards and Shi 1994) using a graphic geometry procedure. Furthermore, if kh = 0
and kv = 0, Eq. (5a) can be reduced to

K pe
K
1 sin 2 Kp
(5b)
K ae 1 sin 1 sin K a

Eq. (5b) is the classical Rankine earth pressure formula for a cohesionless soil with a horizontal
ground surface (Terzaghi 1943, Gnanapragasam 2000), where Ka and Kp are respectively the active
and passive earth pressure coefficients based on Rankines theory as follows

1 sin 1 sin
Kp , Ka (5c)
1 sin 1 sin

3. Distribution of seismic earth pressure

To investigate the distribution of seismic earth pressure, a design example is presented herein.
Consider a rigid retaining wall is constructed to support a c- soil with an infinite top slope in a
seismically active zone. The material properties and seismic parameters are listed in Table 1.

Table 1 Material properties and seismic parameters for a selected earth retaining structure
(kNm-3) c (kPa) () H (m) () kh () kv ()
Unit weight Cohesion Internal friction Height of Slope Horizontal Vertical seismic
of soil of soil angle of soil retaining wall angle seismic coefficient coefficient
18.0 21.6 35 12 10 0.2 0.1
268 Ting-Kai Nian, Bo Liu, Jie Han and Run-Qiu Huang

Calculations were performed using Eq. (3) to obtain the seismic active and passive earth
pressures under horizontal and vertical seismic loading. The distributions of the seismic active and
passive earth pressures along the wall height under upward and toward-the-wall seismic inertia
force are shown in Fig. 2. Fig. 2 (a) shows that the seismic active earth pressure had a triangular
distribution, and a tension crack zone (i.e., the negative seismic active earth pressure) existed
within a critical depth zc from the top of the retaining wall. However, it can be shown from Fig. 2
(b) that the seismic passive earth pressure shows a trapezoidal distribution.

4. Parametric study under seismic active condition

Fig. 2(a) shows a triangular seismic active earth pressure distribution with a tension crack

Active earth pressure AE/kPa

-20 0 20 40 60 80

2 zc
c/H=0.1
4

Depth H/m

6
o
kh=0.2 ()
8 kv=0.1()

10

12

(a) Active condition

Passive earth pressure PE/kPa
0 150 300 450 600 750

2
c/H=0.1
4

Depth H/m

o
6
kh=0.2 ()
8 kv=0.1()

10

12

(b) Passive condition

Fig. 2 Distributions of seismic active and passive earth pressures at the depth from the top of the wall
 Effect of seismic acceleration directions on dynamic earth pressures in retaining structures 269

within the range of critical depth zc. The critical depth zc can be obtained when the seismic active
x
pressure ae = 0. Setting ae 0 in Eq. (3a) with some rearrangements results in a qua-
cos
dratic equation as follows
2
c c
41 AH tan 4 H 2
4 tan 2
2
H H (6)
1 A
2
z z 0
cos
2
1 kv 1 k v 2
By solving Eq. (6), the critical depth zc of the tension crack can be obtained (Nian and Han
2013)
c
2 H
H
sin 1 A 1 A 4 tan 2
2
zc (7)
4 tan 2

1 A 1 k v cos
2

cos 2

Combined with Eq. (3), the active force Pae acting on the retaining wall per unit length can be
expressed as follows (Nian and Han 2013)

1
Pae ae zH H z c 1 K ae zH (1 k v )H H z c (8)
2 2

Eq. (8) can be reformatted in a dimensionless form below (Nian and Han 2013)

z
Pae Pae /( 12 H 2 ) K ae zH (1 k v )1 c (9)
H

To investigate the influence of seismic acceleration directions on the seismic active force on a
retaining wall, four combinations of seismic acceleration with the vertical (upwards or
downwards) and horizontal (towards the wall or backfill) directions were adopted in the present
analysis. The material parameters and seismic parameters used in this analysis are shown in Table
1. A series of dimensionless dynamic active forces were computed by changing the directions of
horizontal and vertical seismic inertia forces, which are shown in Fig. 3.
Fig. 3 shows that the inertia force induced by the horizontal seismic acceleration towards the
wall caused a higher active thrust than that towards the backfill, irrespective of the direction of the
vertical seismic acceleration. This conclusion is also true for a case when only a horizontal seismic
inertia force is applied, which is common in the retaining wall design. However, Fig. 3 shows that
under a downward inertia force with a 0.2 g vertical acceleration, the seismic active thrust Pae had
a maximum value rather than that under an upward inertia force when the horizontal seismic
acceleration was less than 0.35 g. From a practical viewpoint, a greater horizontal thrust Pae
implies a lower factor of safety against sliding and overturning. Therefore, a downward inertia
force should be presented as the proposed approach, and be considered in the design of retaining
structures unless under special circumstances such as a horizontal seismic acceleration higher than
0.35 g.
270 Ting-Kai Nian, Bo Liu, Jie Han and Run-Qiu Huang

0.6
0.5
kv=0.2() c/
0.5 35
o kv=0.5kh()
kv=0


0.4 kv=0.5kh()
kv=0.2()
0.4
2
2PAE/H

0.3

2
2PAE/H
0.3 c/
o
35
0.2

0.2

0.1 0.1

-0.4 -0.2 0.0 0.2 0.4 0.0 0.1 0.2 0.3 0.4 0.5
Towards backfillkh towards wall kh towards wall

Fig. 3 Influence of the direction of inertia force on Fig. 4 Calculated PAE values under downward and
the seismic active thrust upward seismic inertia forces

c/
0.3
0.4 0.00
2
2PAE/H 0.05
0.2 0.10

 0.20
(PAE,KVD-PAE,KVU)/PAE,KVU

c/ 0.3

0.30
0.1

 kv=0.5kh o
0.2 35

0.0
0.1 0.2 0.3 0.4 kh=0.2()
0.1
kh towards wall
-0.1

-1.0 -0.5 0.0 0.5 1.0

-0.2
kv() kv/kh kv()

Fig. 5 Difference of PAE at different soil friction angle Fig. 6 Variation of active force (PAE) with seismic
of coefficient ratio (kv/kh) at different dimen-
sionless cohesion

In the design of retaining structures subjected to seismic loading, the vertical acceleration is
generally taken as half of the corresponding horizontal acceleration according to most earthquake
records (Fang and Chen 1995, Ling and Leshchinsky 1998, Shukla et al. 2009). Fig. 4 shows the
difference in the dimensionless active thrust Pae calculated under upward and downward vertical
accelerations with different horizontal seismic accelerations. The value of Pae, calculated under a
downward seismic acceleration was higher than that under an upward seismic acceleration when
the horizontal seismic acceleration was less than 0.35 g. However, a different conclusion can be
made when the horizontal seismic acceleration was larger than 0.35g, i.e., Pae calculated under a
downward seismic acceleration was lower than that under an upward seismic acceleration.
Fig. 5 shows the difference in dimensionless active thrust Pae determined by the vertically
downward and upward seismic accelerations for soil backfills with friction angles of 30 to 40.
The subscripts KVD and KVU indicate that kv acted downwards and upwards respectively. The
 Effect of seismic acceleration directions on dynamic earth pressures in retaining structures 271

maximum difference in the calculated seismic active thrust under upward and downward seismic
inertia forces reached 25%. Fig. 5 also shows that the upward seismic acceleration resulted in a
higher Pae when kh exceeded 0.35. In fact, an earthquake with a horizontal acceleration greater than
0.35g is not very common (Fang and Chen 1995, Ling and Leshchinsky 1998). However, a correct
acceleration direction should be chosen to ensure the safety of designed retaining structures
located within a seismic active zone.
At various dimensionless cohesion c/ H, friction anlges of soils, and slope angles , and a
given horizontal seismic inertia force towards the wall (kh = 0.2), the relationships between the
seismic active force Pae and the seismic coefficient ratio kv/kh, were established and are shown in
Figs. 6 to 8. Figs. 6 to 8 show that the active forces increased linearly with the increase of the
seismic coefficient ratio from 1.0 to +1.0, regardless of the direction of the vertical seismic inertia
force. Figs. 6 and 7 shows that a higher seismic active force was obtained when the vertical
seismic inertia force acted downwards no matter whether the soil properties, such as dimensionless
cohesion c/ H, friction angle , and slope angle changed. Under the given horizontal and vertical
seismic inertia forces, the seismic active force Pae increased with the increase of the slope angle
(Fig. 8) while it decreased with the increase of the dimensionless cohesion c/ H (Fig. 6) or the soil
friction angle (Fig. 7).
The critical depth zc of a tension crack behind the wall can be computed using Eq. (7). Fig. 9
shows the relationship between the relative critical depth zc/H and the horizontal seismic
acceleration coefficient kh (towards the wall) under an upward or downward seismic inertia force.
It is shown that the relative critical depth zc/H decreased with the increase of the horizontal seismic
coefficients kh. At the given soil properties and the horizontal seismic coefficient, a greater critical
depth of the tension crack was observed under an upward seismic inertia force. Fig. 10 shows the
curves of relative critical depth zc/H and dimensionless cohesion c/ H at the given soil properties
and the horizontal and vertical seismic inertia forces. It is shown that the relative critical depth
zc/H increased linearly with the increase of the dimensionless cohesion c/ H. Fig. 11 shows the
nonlinear decrease of the active force with the cohesion.

c/ c/
o
0.00 35 0.25
0.4
2

0.05 kh=0.2()
2PAE/H

0.10 0.20
0.3 0.20
0.30
0.15
o
0.2 35

0.10
2

2PAE/H

kh=0.2()
0.1

0.05

-1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0
kv() kv/kh kv() kv() kv/kh kv()

Fig. 7 Variation of active force (PAE) with seismic Fig. 8 Variation of active force (Pae) with seismic
coefficient ratio (kv/kh) at different soil coefficient ratio (kv/kh) at different slope
friction angle angle
272 Ting-Kai Nian, Bo Liu, Jie Han and Run-Qiu Huang

0.40 1.0
kv=kh(
0.35 kv=0.5kh(
0.8 kv=0
0.30
kv=0.5kh(
0.25 0.6
zc/H kv=kh(
kv=kh(
0.20 zc/H
kv=0.5kh(
0.15 0.4
kv=0
o
kv=0.5kh( 35 o
0.10 35
c/ 0.2
kv=kh(

0.05
kh=0.2()
0.00 0.0
0.0 0.1 0.2 0.3 0.4 0.00 0.05 0.10 0.15 0.20 0.25 0.30
kh towards wall c/H

Fig. 9 Variation of critical depth (zc/H) with Fig. 10 Variation of relative critical depth (zc/H) with
horizontal seismic coefficient (kh) under dimensionless cohesion c/ H under horizontal
upward or downward inertia force (kv) and vertical inertia forces

0.5
6

2
o
35

2PPE/H

kv=kh(
0.4 5 c/H=0.1
kv=0.5kh(
kv=0 4
0.3
kv=0.5kh(
2
2PAE/H

kv=kh( 3
kh(), kv=0 kh(), kv=0
0.2
kh(), kv=kh() 2 kh(), kv=kh()
o
35 kh(), kv=kh() kh(), kv=kh()
0.1

1
kh(), kv=0.5kh() kh(), kv=0.5kh()
kh=0.2()
kh(), kv=0.5kh() kh(), kv=0.5kh()
0.0 -0.4 -0.2 0.0 0.2 0.4
0.05 0.10 0.15 0.20 0.25 0.30
c/H Towards backfillkh towards wall

Fig. 11 Variation of active force (Pae) with c/ H under Fig. 12 Influence of seismic inertia force direc-
horizontal and vertical inertia forces tions on passive resistance

5. Parametric study under seismic passive condition

Fig. 2 (b) shows a trapezoid seismic passive earth pressure distribution. The seismic passive
force Ppe acting on the rigid retaining wall per unit length can be expressed below by combining
Eqs. (3) and (4)
sin 1
Ppe
1
2

pe z 0 pe zH H 12 K pe z H (1 k v )H 2c
cos

cos H
(10)
1 sin 1
z H (1 k v )H cH cos
2
K pe
2 cos
The above equation can be expressed in a dimensionless form as follows
 Effect of seismic acceleration directions on dynamic earth pressures in retaining structures 273

kh

2
2PPE/H
0.0 0.1 0.2 0.3 0.4 6

c/ 5

(PPE,KHW-PPE,KHB)/PPE,KHB

-0.1
kvkh() 4
-0.2
3


c/
-0.3 0.00

2
35
o 0.05

0.10
-0.4 1
0.20
kh=0.2()
0.30
-0.5 -1.0 -0.5 0.0 0.5 1.0
kv() kv/kh kv()

Fig. 13 Difference of PPE at different friction angle Fig. 14 Variation of passive force (Ppe) with
of using the proposed and conventional seismic coefficient ratio (kv/kh) at diffe-
methods rent dimensionaless cohesion

2PPE/H
5
5


4
2
2PPE/H

3
3

2
2
c/
c/ o

35
1 1

kh=0.2()
kh=0.2

-1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0
kv() kv/kh kv() kv() kv/kh kv()

Fig. 15 Variation of passive force (Ppe) with seis- Fig. 16 Variation of passive force (Ppe) with seis-
mic coefficient ratio (kv/kh) at different mic coefficient ratio (kv/kh) at different
friction angle slope angle

c 1 sin
Ppe Ppe /( 12 H 2 ) K pe zH (1 k v ) 2 cos (11)
H cos

To investigate the influence of the seismic acceleration directions on the seismic passive forces,
the material parameters and seismic parameters shown in Table 1 were adopted with Eq. (11). A
series of dimensionless seismic passive forces were obtained and are shown in Fig. 12 by changing
the directions of horizontal (towards the wall or backfill) and vertical (upwards or downwards)
seismic inertia forces. From a practical viewpoint, the minimum passive force should be
determined for resisting the failure of a retaining structure. Fig. 12 shows that a minimum passive
force Ppe was caused by the upward and towards-the-wall inertia forces based on the solution
274 Ting-Kai Nian, Bo Liu, Jie Han and Run-Qiu Huang

proposed in this study. This result is different from Kapilas finding that the critical inertia forces
acted upwards and towards the backfill using the conventional passive force equation (Kapila
1962). The differences in the calculated dimensionless passive resistance Ppe using the present and
conventional (i.e., by Kapila in 1962) methods for the soil friction angles ranging from 30 to
40 are shown in Fig. 13. The subscripts KHW and KHB denote the inertia forces with kh acting
towards the wall in the present method and the backfill in the conventional method respectively.
The maximum difference in the seismic passive resistance between the toward-the-wall and
toward-the-backfill seismic inertia forces was approximately 50%. Fig. 13 also shows that the
difference became greater when the horizontal seismic coefficient increased.
Figs. 14 to 16 show the relationships between the seismic passive force Ppe and the seismic
coefficient ratio kv/kh at a given horizontal seismic inertia force towards the wall (kh = 0.2) but at
different dimensionless cohesion c/ H, friction angle and slope angle . It is shown that at a
given horizontal seismic inertia force towards the wall, the passive force increased linearly with
the increase of the seismic coefficient ratio from 1.0 to +1.0, regardless of the direction of the
vertical seismic inertia force. Moreover, the lower seismic passive force was obtained when the
vertical seismic inertia force acted upwards, irrespective of the change of the parameters such as
c/ H (Fig. 14), (Fig. 15), and (Fig. 16). At the given horizontal and vertical seismic inertia
forces, the passive force Ppe decreased with the increase of the slope angle (Fig. 16) while that
increased with the increase of the dimensionless cohesion c/ H (Fig. 14) or friction angle (Fig.
15).

6. Discussions

Several analytical approaches are frequently used in current practice to calculate the seismic or
dynamic earth pressures/forces on retaining structures, such as the elastic theory (Wood, 1973),
Coulombs sliding wedge theory (Mononobe 1924, Okabe 1924, Seed and Whitman 1970) and
Rankines limit stress state theory (Terzaghi 1943, Richards and Shi 1994). However, there exist
some significant differences in the solutions from different methods because these methods are
used for specific applications in the design of retaining structures subjected to seismic excitation.
For example, Wood (1973) assumed a linear elastic behavior of the wall-soil. Compared with the
linearly elastic assumption made by Wood (1973), the commonly used Mononobe-Okabe method
based on Coulombs sliding wedge theory employs the assumption of sufficiently large wall
deformations to induce a fully plastic stress condition in the soil, while the limit stress state
approach based on Rankines earth pressure theory allows a tensile crack at the top of the wall. In
order to illustrate the differences in the several seismic earth pressure/force approaches, a typical
rigid retaining structure under a horizontal seismic acceleration with ah = 0.15 g is employed, and
the wall is assumed to be vertical and smooth, while the fill consists of sandy soils with c = 0, =
36and = 0.3. Comparison of the solutions from different methods are shown in Figs. 17 and 18,
respectively, in which the parameter Pae represents the active wall force increment due to
horizontal earthquake loading, and the other parameters are defined earlier.
It can be shown from Fig. 17 that the results from the improved Rankines earth pressure theory
proposed in this paper fall in between those based on Coulombs earth pressure theory such as the
Mononobe-Okabe and Seed-Whitman approaches. Moreover, the seismic force increments in this
study are generally close to those by the Mononobe-Okabe approach under a high friction angle,
and tend to close to those by the Seed-Whitman approach under a low friction angle. Particularly,
 Effect of seismic acceleration directions on dynamic earth pressures in retaining structures 275

the results from the present study and Coulombs sliding wedge concept are far lower than those
1.0 1.0
Wood Wood

0.8 0.8
dy
stu
is
Th

PAE/[H (ah/g)]
PAE/[H (ah/g)]

0.6 0.6

2
2

Seed & Whitman ( kv=0) Seed & Whitman ( kv=0)

0.4 0.4
This study
0.2 Mononobe-Okabe 0.2 Mononobe-Okabe
o
( =36 )
(ah=0.15g)

0.0 0.0
28 30 32 34 36 38 40 42 44 0.0 0.1 0.2 0.3 0.4 0.5
deg ah (g)

Fig. 17 Variation of seismic force increments with Fig. 18 Variation of seismic force increments with
friction angle at ah = 0.15 g horizontal seismic accelerations at = 36

from Woods solution.

Fig. 18 shows that the results in this study are generally closer to those by the
Mononobe-Okabe and Seed-Whitman approaches under low-moderate horizontal seismic
accelerations, while they are close to those by Woods elastic theory under a high horizontal
seismic acceleration. That is to say, the proposed approach has a wide application for different
horizontal seismic acceleration coefficients. According to the fact that very limited wall failures
were observed during strong earthquakes, thus the seismic force increments from Woods elastic
theory are remarkably higher than the practical values that the retaining structures undergo.
Therefore, care is required in selecting the most appropriate method to calculate the seismic forces
on retaining structures under a particular seismic situation.

7. Conclusions

Based on the development of the theoretical solution and the discussion on the influence factors
on seismic active and passive forces on retaining structures under seismic loading, the following
conclusions can be made:
The present study developed an analytical solution for the active and passive forces on a
rigid retaining structure with c- backfill and an infinite top slope considering both the
horizontal (towards the wall or the backfill) and vertical (upwards or downwards) seismic
inertia forces based on Rankines earth pressure theory and the Mohr-Coulomb yield
criterion.
For a retaining structure with c- backfill under a seismic active condition, the critical
inertia forces to cause a maximum active thrust Pae should act downwards and towards the
wall when the horizontal seismic acceleration is less than 0.35 g. When the horizontal
seismic acceleration is larger than 0.35 g, however, the critical inertia forces to cause a
maximum active thrust Pae should act upwards and towards the wall. Therefore, it is
important to evaluate the vertical seismic inertia force in a correct direction during the
276 Ting-Kai Nian, Bo Liu, Jie Han and Run-Qiu Huang

design of a retaining structure under seismic loading.

For a retaining structure under a seismic passive condition, the critical inertia forces to cause
a minimum passive resistance Ppe should act upwards and towards the wall.
When only a horizontal seismic inertia force is considered in the design of a retaining
structure, it is also true that the inertia force towards the wall causes a higher active thrust or
a lower passive resistance than that towards the backfill.
A tension crack is found behind the retaining structure with a c- backfill under seismic
loading. A greater critical depth of the tension crack is observed under an upward seismic
inertia force. The relative critical depth zc/H decreases linearly with the increase of the
horizontal seismic inertia force and increases with the increase of the dimensionless
cohesion c/ H and thus leading to the decrease of the active force.

Acknowledgments

The authors wish to acknowledge the support of the National Natural Science Foundation of
China (Grant No.51179022) and the Open Research Fund of the State Key Laboratory of
Geohazard Prevention and Geoenvironmental Protection, Chengdu University of Technology
(Grant No. SKLGP2010K005) for this research. In addition, the authors appreciate the anonymous
reviewers for their helpful comments, which have helped the quality of the paper.

References
Budhu, M. and Al-Karni, A.V. (1993), Seismic bearing capacity of soils, Geotechnique, 43(1), 181-187.
Chen, W.F. (2007), Limit Analysis and Soil Plasticity, J. Ross Publishing, Fort Lauderdale, FL, USA.
Choudhury, D. and Nimbalkar, S. (2005), Seismic passive resistance by pseudo-dynamic method,
Geotechnique, 55(9), 699-702.
Das, B.M. (2008), Fundamentals of Geotechnical Engineering, (3rd Edition), Cengage Learning, Stanford,
CA, USA.
Fang, Y.S. and Chen, T.J. (1995), Modification of Mononobe-Okabe theory, Geotechnique, 45(1), 165-
167.
Ghosh, P. (2008), Seismic active earth pressure behind a nonvertical retaining wall using pseudo-dynamic
analysis, Can. Geotech. J., 45(7), 117-123.
Gnanapragasam, N. (2000), Active earth pressure in cohesive soils with an inclined ground surface, Can.
Geotech. J., 37(2), 171-177.
Kapila, J.P. (1962), Earthquake resistant design of retaining walls, Proceedings of the Second Earthquake
Symposium, Roorkee, India, December.
Lancellotta, R. (2002), Analytical solution of passive earth pressure, Geotechnique, 52(8), 617-619.
Lancellotta, R. (2007), Lower-bound approach for seismic passive earth resistance, Geotechnique, 57(3),
319-321.
Ling, H.I. and Leshchinsky, D. (1998), Effects of vertical acceleration on seismic design of geosynthetic-
reinforced soil structures, Geotechnique, 48(3), 347-373.
Mononobe, N. (1924), Consideration into earthquake vibrations and vibration theories, J. Japan. Soc.
Civil Eng., 10(5), 1063-1094.
Nian, T.K. and Han, J. (2013), Analytical solution for seismic earth pressures in c- soil with an infinite
slope, Technical Note, ASCE J. Geotech. Geoenviron. Eng., 139(9), 1611-1616.
Okabe, S. (1924), General theory on earth pressure and seismic stability of retaining wall and dam, J.
Japan. Soc. Civil Eng., 10(5), 1277-1323.
Richards, R. and Shi, X. (1994), Seismic lateral pressures in soils with cohesion, J. Geotech. Eng., ASCE,
 Effect of seismic acceleration directions on dynamic earth pressures in retaining structures 277

120(7), 1230-1251.
Richards, R., Elms, D.G. and Budhu, M. (1990), Dynamic fluidization of soils, J. Geotech. Eng., ASCE,
116(5), 740-759.
Seed, H.B. and Whitman, R.V. (1970), Design of earth retaining structures for dynamic loads,
Proceedings of Special Conference on Lateral Stresses in the Ground and Design of Retaining Structures,
Ithaca, New York, USA, June.
Shukla, S.K. and Habibi, D. (2011), Dynamic passive pressure from c- soil backfills, Soil Dyn. Earthq.
Eng., 31(6), 845-848.
Shukla, S.K., Gupta, S.K. and Sivakugan, N. (2009), Active earth pressure on retaining wall for c- soil
backfill under seismic loading condition, J. Geotech.Geoenviron. Eng., ASCE, 135(5), 690-696.
Terzaghi, K. (1943), Theoretical Soil Mechanics, John Wiley Sons, Inc., New York, NY, USA.
Wood, J.H. (1973), Earthquake-induced soil pressures on structures, Ph.D. Dissertation, California
Institute of Technology, Pasadena, CA, USA.
Yao, L., Feng, J. and Yang, M. (2009), Damage analysis of subgrade structures in Wenchuan earthquake
and recommendations for improving seismic design code, J. Southwest Jiaotong Univ., 44(3), 301-311.

GC