Proceedings of Indian Geotechnical Conference

December 15-17,2011, Kochi (Invited Talk -5.)

DYNAMIC EARTH PRESSURE AGAINST RETAINING WALLS

Vijay K. Puri, Professor, Civil and Environmental Engineering,SIU Carbondale, IL, USA , [email protected]
Shamsher Prakash, Professor Emeritus, Civil Engineering, MUST Rolla, MO, USA, [email protected]

ABSTRACT.. The paper presents a simple method for calculation of static and dynamic active force on the rigid retaining
wall following the pseudo-static approach of analysis and includes the effects of cohesion of the backfill and the friction
between the backfill and the wall face. The resultant earth pressure obtained by this method has been compared with the
experimentally observed values during small scale tests on retaining walls by other investigators. The displacement of retaining
wall must not exceed specified values. The methods for calculation of displacement of rigid retaining walls are also discussed.

INTRODUCTION
There are two aspects in the design of retaining walls for
q C D
earthquake loading:
(a). Calculation of static and dynamic lateral pressures. A Zone of tensile
i E crack
(b). Calculation of displacement of the retaining wall. crack
Hc
The lateral earth pressure for static case is generally C
determined using either Rankine’s [1] or Coulomb’s [2] Ca IF C
method. The earthquake induced forces on the retaining wall W
Wr
are generally computed using the modified Coulomb’s H G
approach in which the earthquake force on the backfill is
replaced by an equivalent static force. This is known as Pa(dyn R

Mononobe-Okabe method [3,4]. A solution for determination B

of static and dynamic active earth pressure for c- soils was Fig. 1 Typical assumed failure wedge with forces acting on it.
developed earlier by Saran and Prakash [5]. This provides a
convenient method for determination of static and dynamic
lateral pressures for a typical soil. However, it has the
following limitations: c’a
1. The effect of the vertical component of acceleration been
neglected. c
2. The backfill surface is assumed to be horizontal, which (Pa)dyn
may actually be inclined in many cases. (W+Q)(1+ v)
3. The unit adhesion between the back face of the retaining
wall and soil was assumed to be equal to the unit cohesion of R
the soil
A method for calculation of dynamic active pressure for a c-
soil accounting for the effect of the following factors is IF=(W+q) h
presented here(Fig. 1). Figure 2. Polygon of Forces
The effect of cohesion, c, and adhesion, ca.
The inclination of the backfill, i The polygon of forces acting on the assumed failure wedge
Horizontal and vertical seismic coefficients, h and v may then be drawn as shown in Fig.2. Applying the
respectively. conditions of equilibrium and simplifying the results a
Surcharge, q relationship for (Pa)dyn can be obtained as
Inclination of the wall face,
( Pa ) dyn 1 H 2 (N
Depth of tension cracks, Hc 2 a ) dyn

EARTH PRESSURE CACULATION qH ( N aq ) dyn cH ( N ac ) dyn

………….…(1)
A schematic diagram of dynamic earth pressure problem is
in which,
shown in Fig.1. ABEC is an assumed failure wedge.
Considering the unit length of the wall, (Pa)dyn = total active cos( i ) cos e cos(
( N ac ) dyn [ ]
force, R = soil reaction, IF = horizontal inertia force, W = cos cos( i) cos
weight of assumed failure wedge, Wt=resultant of weight W
and IF, C= cohesion force, Ca =adhesive force, q = surcharge, 1
x
= inclination of wall face with vertical ( > 0) and i = sin( ) …………………….…...(2)
inclination of the backfill ( < i < ).

29
Puri, VK and Prakash,S

sin( )
( N aq ) dyn [ n tan cos ] x 0.8
cos
=0, i = 0, v =0
0.7
sin( ) (1 v ) cos( )
[ h
sin( ) cos( i) ………...……(3) 0.6

1
( N a ) dyn tan tan 0.5
sin( )
n

KAE cos
[( 2 n ) tan cos 2 sin ] 0.4
cos i cos( i)
2
sin ( ) sin i 0.3
}
2
cos cos cos( i) Author’s result
0.2
[ h sin( ) (1 v ) cos( )] Mononobe-Okabe
…………(4) theory
and
0.1 Model test results [6]
ca
e ……………………………… (5)
c
0
Where ca = unit adhesion between the back face of the wall
0 0.1 0.2 0.3 0.4 0.5 0.6
and the backfill, and c = unit soil cohesion.
h

The static earth pressure can be obtained from Eqs. 1-5, by Fig.3. Translation of wall for active pressure – comparison of
making h and v as zero. The value of (Pa)dyn and (Pa)stat theory with model test results
obtained from Eq. (1) are for the assumed failure wedge. In
order to obtain the maximum values of the total dynamic computed by the proposed method show an excellent
earth force, (Pa)dyn , the earth pressure coefficients (Naq)dyn, agreement with Mononobe-Okabes method and are
(Na )dyn, and (Nac)dyn were optimized. A computer code was somewhat smaller than the those observed by sheriff et al[6].
developed for this purpose. It must be mentioned here that The effect of various parameters on dynamic active earth
these earth pressuree coefficients were individually optimized pressure was investigated using equation 1. It was observed
and then (Pa)dyn was obtained by superimposing their effect, that a conservative value of dynamic active earth pressure can
i.e., using Eq. (1). The same procedure was followed for the be obtained by assuming ca=c.
maximum value of static earth force, (Pa)stat. From known
values of (Pa)dyn and (Pa)stat, the dynamic increment ( Pa)dyn POINT OF APPLICATION
can be obtained as Mononobe-Okabe’s solution had assumed that the resultant
( Pa)dyn = (Pa)dyn – (Pa)stat………………...……………...(6) active thrust acts at a distance of H/3 from the bottom of the
wall similar to the static case ( h = v =0). The laboratory
COMPARISON WITH OBSERVED DATA observations indicate that the resulting active thrust acts
Sherif, et al. [6] reported results of measurements of somewhat higher than H/3 measured from the bottom of the
dynamic active earth pressure on a 1 m high rigid retaining wall. Seed and Whitman [7] have suggested that for the case
wall. The backfill properties are as follows: of rotation about the bottom of the wall, the static pressure
Unit weight, = 16.28 kN/m3 may be assumed to act at H/3 and the dynamic increment at
Angle of internal friction, = 40.9o 0.6 H from the base of the wall. For a wall undergoing
Angle of wall friction = 23.9o rotation about the top, the resulting active thrust may be
Ssope of backfill = 0 o assumed to act at 0.55H from the bottom of the wall [8]. It
The wall was subjected to sinusoidal acceleration of up to 0.5 has been suggested that for wall undergoing translation, the
g. The results are shown in Fig..3. It may be noted that line of the static active thrust may be assumed to act at 0.42
( Pa ) dyn H and dynamic increment at 0.48 H above the bottom of the
K AE ………………………………(7) wall [7]. For all these case the active thrust and dynamic
1 H2 increment are assumed to act at angle ’ ’ with the normal to
2
The result obtained from Mononobe-Okabe theory and from the wall face.
the present study are also shown in Fig.3. It may be observed
from Fig. 3 that the computed dynamic active earth pressures

30