Chapter 2. Lateral Earth Pressure
Chapter 2. Lateral Earth Pressure
Chapter 2. Lateral Earth Pressure
Fig. 2.1 a
Fig. 2.1 b
31
where
G
G
PG
sin ( )
sin ( )
G S ABC 1 f ( H , , , )
P f1 ( , H , , , , , )
(2.1)
In relation (2.1) the only variable is . In order to find the maximum value of
P, corresponding to a particular value of , the first derivate
should be
(2.2)
(2.3)
1
H 2 Ka
2
(2.4)
32
, , ,
tan 2 (45 o
),
H 2 tan 2 (45 o )
2
2
(2.5)
Fig. 2.2
Another line BE, called orientation line, at an angle to the BD line is drawn.
Trial failure surface BC1, BC2, BC3... are considered.
The weight G1 of the wedge ABC1 is computed and represented at a chosen
scale on the reference line BD. From the extremity of the vector G 1, a parallel
to the orientation line BE is drawn. The parallel intersects the failure plane
BC1 in the point P1. The vector OP1 represents the reaction P1 corresponding
to the wedge of soil ABC1.
33
One should check that the triangle BG 1P1 in the Culmanns construction is
equal to the triangle of forces G 1, P1, R1, but rotated at an angle (90 o + ) to
the vertical. The side BG 1 is equal to the weight G 1. The two angles adjacent
to the side BG1 are: BG1 P1 = and P1BG1 = ( ) . The two triangles are
equal.
The procedure is repeated for the wedges ABC 2, ABC3, ABC4, giving the
corresponding forces P2, P3, P4. The extremities of the vectors P 1, P2, P3
are linked with a continuous curve. A tangency to the curve parallel to the
reference line defines the maximum value P max = Pa.
The direction o of the failure plane is obtained by linking the point of
tangency with the point B.
Active earth pressure distribution
The analytical approach by Coulomb or its graphical transposition by
Culmann lead to the determination of the total active thrust Pa.
In practice, however, is necessary to know also the distribution of the active
earth pressure on the wall.
In the case of a plane soil surface, it is commonly accepted a linear
distribution with depth of the earth pressure (fig. 2.3).
Fig. 2.3
The area of the triangle of pressures should be equal to the force P a which
was established analytically or graphically. From that condition, the ordinate
of the pressure diagram at the base, paH, is found.
1
p aH AE Pa
2
AE = AB
AE AB cos
1
cos
p aH H
Pa
2
sin
p aH
H
cos
sin
2 Pa sin
H cos
(2.6)
90 o , 0
34
p aH
2 Pa
H
(2.7)
where:
G
G
G S ABC 1 f ( H , , , )
35
P f1 ( , H , , , , , )
P
0
(2.8)
(2.9)
1
H2 Kp
2
(2.10)
, , ,
For the particular case: 90; 0; 0; K p tan 2 ( 45o 2 ) value which was
H 2 tan 2 (45 o )
2
2
(2.11)
36
Fig. 2.5
The extremities of the vectors P1, P2, P3 are linked with a continuous curve.
A tangency to the curve parallel to the reference line defines the minimum
value Pmin = Pp.
The direction o of the failure plane is obtained by linking the point of
tangency with the point B.
Passive earth resistance distribution
As for the active case, it is assumed a linear distribution with depth of the
passive earth resistance (fig. 2.6).
Fig. 2.6
The area of the triangle of passive earth resistances should be equal to the
force Pp which was established analytically or graphically. From that condition
the ordinate of the passive earth resistances diagram at the base, p pH, is
found.
1
p pH AE Pp
2
37
AE AB cos
H
cos
sin
1
cos
p pH H
Pp
2
sin
2 Pp sin
p pH
H cos
(2.12)
90 o , 0
2 Pp
H
(2.13)
tan c
(2.14)
(2.15)
For a given failure plan, the cohesion force C is known both in magnitude and
direction. The equilibrium of the soil wedge will imply to decompose the
resultant of forces G and C on the directions of P and R (fig. 2.8). A number of
trial failure planes would have to be selected and polygons of forces similar to
the one in fig. 2.8 to be constructed. In this way, the maximum active thrust P a
is obtained (fig. 2.9).
Fig. 2.8
Fig. 2.9
A more refined computation will consider also the influence of the adhesion
force Ca developed along the wall and the fact that cracks might develop
within the tension zone at a depth
zo
2c
tan ( 45 o
).
1
2
w zo
2
(2.16)
(2.17)
There are two forces for which only the direction is known: P and R. They are
found from the polygon of forces expressing the equilibrium condition (fig.
2.10).
The construction is repeated for other trial failure planes BC 2, BC3etc, to
find the failure surface for which Pmax = Pa (fig. 2.11).
39
Fig. 2.10
Fig. 2.11
AC
is added:
1
AB CC ' q AC
2
CC ' AC sin( )
AB
H
sin
Gtotal
sin ( )
sin ( )
2q
sin
1
1
H AC
q AC H AC
1
2
sin
2
sin
H sin ( )
Gtotal
H
2H e
sin ( )
1
H AC
(1 2 e ) G (1
)
2
sin
H
H
where
He
sin
sin ( )
q
(2.18)
(2.19)
40
Fig. 2.13
In the triangle A A A: He = A A = A ' A sin
'
A ' A '' h
'''
'
'''
H e A ' A sin
sin
sin ( )
q
, 0
41
He h
He = h =
(2.20)
He
H
triangle of forces will include Gtotal Ptotal R, where Paq is the total active thrust
including the effect of the surcharge q. The sine rule is applied, leading to:
Paq
sin ( )
Paq Gtotal
By derivation,
Gtotal
sin ( )
2H e
sin ( )
sin ( )
G
(1
)
sin ( )
sin )
H
Paq
0,
the term ( 1
2H e
H
(2.21)
is preserved. Therefore, Paq and P will be in the same relation as Gtotal and G:
Paq Pa (1
2H e
)
H
(2.22)
42
Fig. 2.14
Particular case: 90 ,
o
0, 0
(fig. 2.15)
H 2 tan 2 ( 45 o ) H H e tan 2 ( 45 o )
2
2
2
2H e
2H e
1
H 2 tan 2 ( 45 o ) 1
Pa (1
)
2
2
H
H
Paq Pa Pa
(2.23)
Fig. 2.15
A similar procedure is used in the case of the passive resistance. For
instance, when 90 o , 0 (fig. 2.16)
1
H 2 tan 2 ( 45 o ) 2 H H e tan 2 ( 45 o )
2
2
2
2
H
2
H
1
e
e
H 2 tan 2 ( 45 o ) (1
) Pp (1
)
2
2
H
He
Ppq Pp Pp
(2.24)
Fig. 2.16
The case of stratified soil mass
43
The concept of equivalent height is used also to compute the earth pressure
in the case of a stratified soil mass.
In fig. 2.17 is shown a vertical wall limited by a horizontal surface on which is
acting a uniformly distributed surcharge pressure q. There are two layers of
soil behind the wall, with characteristics 1 , 1 and, respectively 2 , 2 .
Fig. 2.17
The following approximate method is applied in order to compute the active
earth pressure diagram.
Computation starts with the top layer 1. The surcharge pressure q is replaced
q
1
by layer of soil with unit weight 1 and height he . The ordinates of the
1
earth pressure diagram at the top and at the base of the layer 1 are then
computed:
1
1
)
2
p abase 1 ( he h1 ) tan 2 ( 45 o
1
(2.25)
1
)
2
(2.26)
Passing to the layer 2, the vertical pressure of the layer 1, including the
fictitious layer of height h e1 replacing the surcharge pressure q, represents a
surcharge pressure for the layer 2. According to the known rule, this must be
replaced by a layer of soil with unit weight 2 and equivalent height
he
2
1 ( he h1 )
2
1
(2.27)
44
The ordinates of the earth pressure diagram at the top and at the base of the
layer 2 are computed.
p atop 2 he tan 2 (45 o
2
2
)
2
(2.28)
2
)
2
(2.29)
If the soil would have been formed by n layers, the relations for the n layer
are:
p atop n he tan 2 ( 45 o
n
n
)
2
p abase n ( he hn ) tan 2 ( 45 o
n
(2.30)
n
)
2
(2.31)
As one can realize, the earth pressures diagram has jumps at the lines of
separation between layers. The jumps occur when the angles of internal
friction of various layers are different. If the angle would have been
constant and only the unit weights different, the diagram would present
only changes in slope from one layer to the next one.
In the general case of a wall with inclined face and sloping surface of soil, the
Coulombs method is used, starting with the upper layer, transformed then in
a surcharge pressure for the next layer a.s.o.
The earth pressures diagrams with jumps at the change of the layer are
considered as conventional ones because admit than in the same points of
base
top
the wall are acting simultaneously two pressures ( pa and pa a.s.o.). In
reality, the earth pressures diagram should be a continuous one, without
jumps. Nevertheless, for practical purpose the procedure described is
accepted.
1
45
Fig. 2.18
To establish the magnitude of the total active thrust PaQ, including the effect of
the surcharge Q, the graphical method of Culmann is used.
Trial failure planes AC1, AC2...etc are considered in turn and the weights of
the soil wedges ABC1, ABC2, including, when occurs, the surcharge Q, are
plotted on the reference line. Culmanns curve has a jump when the failure
plane passes through the point of application of the surcharge Q. In most
cases, this gives the maximum value of the thrust.
PaQ = Pa + Pa
(2.32)
There are, however, situations when the failure plane does not pass through
the point of application of the surcharge Q or when the surcharge is so far in
respect to the crest A of the wall that it does not influence the magnitude of
the active thrust.
Influence of the position of the surcharge Q on the magnitude of the active
thrust
Two Culmanns curves are drawn (fig. 2.19).
The curve 1 corresponds to the case when the surcharge Q is considered to
act from the very first wedge.
The curve 2 corresponds to the case where there is no surcharge.
Except the initial part, the two curves are parallel.
Tangents are drawn to each of the two curves, by means of which three
zones are defined on the soil surface.
46
Fig. 2.19
Zone I
AC1 The point C1 is obtained by linking B with the point of tangency to the
curve 1. PaQ is the maximum total active thrust. No matter where Q is located
inside this zone, the additional thrust P ( P P ) is constant.
a
aq
Zone II
C1C2 The tangency to the curve 2 intersects the curve 1 in the point M.
Linking the point B with the point M and extending the line BM, the point C 2 is
defined on the surface. When the surcharge Q acts in the zone II, the failure
plane is passing through the point of application of the surcharge. The
additional active thrust Pa varies between a maximum, corresponding to the
point C1 and zero, corresponding to the point C2.
Zone III
To the right of C2. For a given failure plane BC; one can realize that the thrust
P 'a , given by the intersection with the curve 1 is smaller than the thrust P a
corresponding to the curve 2, when there is no surcharge. In conclusion, for
any point to the right of C 2, the surcharge does not influence the magnitude of
the active thrust.
The variation of the additional thrust Pa in the zones I and II is shown in the
fig. 2.20.
47
Fig. 2.20
Changes of the active earth pressures diagram
a. The surcharge Q acts in zone I (fig. 2.21)
Fig. 2.21
From the point of application of the surcharge Q a parallel is drawn to the line
of natural slope, defining the point M and another parallel is drawn to the
failure line BC1, defining the point N. Changes in the pressures diagram occur
between the ordinates corresponding to the points M and N. The area of the
triangle of additional pressures is equal to Pa.
b. The surcharge acts in zone II (fig. 2.22)
The failure plane passes through the point of application of the surcharge Q.
A parallel is drawn from this point to the line of natural slope, defining the
point M. Changes in the pressures diagram occur below the ordinate
corresponding to the point M. The area of the triangle of additional pressures
is Pa.
48
Fig. 2.22
Both diagrams shown in the fig. 2.21 and 2.22 are conventional ones, since
they admit jumps in the point M but, nevertheless, they are accepted in the
practice.
2.1.6 Earth pressure in cases with particular geometry
a. Face of the wall with two slopes (fig. 2.23a)
The Culmanns procedure is used. Computation starts with the face AB 1,
considered as a distinct wall. The total active thrust Pa 1 is obtained by
drawing the tangency to the Culmanns curve parallel to the reference line.
The corresponding pressure diagram on the face AB 1 is then computed (fig.
2.23b).
The face B1B2 of the wall is considered as part of a fictitious wall with the face
A ' B1B2. The Culmanns construction is used for this wall, the total active
thrust on the face A ' B1B2 and the corresponding pressure diagram are
obtained. From the total pressure diagram on the fictitious wall A ' B1B2 is then
considered only the part corresponding to the real face B 1B2 in order to define
Pa2.
b. Ground surface with two slopes (fig. 2.24)
The magnitude of the total active thrust can be computed by replacing the
wall AB with a fictitious wall B A ' chosen in such a way that the weight of the
failure wedge remains unchanged. For that purpose, the point B is linked to
the point C, where the change in the slope of the ground surface occurs, and
then a parallel to BC is drawn from A, until it reaches in A ' the extension of
the slope CD.
Fig. 2.24
A ' B fulfills the required condition: S ABC S A ' BC because the two triangles
have the same base (BC) and the same height h (the distance between the
parallels A A ' and BC).
The total active thrust P a on the fictitious wall AB and one slope ground
surface A ' CD can be found using Culmanns method, with the observation
that the orientation line should consider the inclination of the face AB.
To determine the earth pressure diagram, the one-slope ground surface ACM
is considered and the Culmanns method is used, thus obtaining the failure
surface BC1 and the pressure diagram abc. The total active thrust Pa
corresponding to the wall AB limited by the one-slope ground surface ACM is
1
50
2 ( Pa Pa )
1
(2.33)
The total active thrust on the wall AB, limited by the two-slopes ground
surface is the area of pressures abpn and is acting in the centre of gravity of
that area.
2.2 The influence of the wall-soil friction and of the deformation
condition of the wall on the lateral earth pressure
2.2.1 Soil-wall friction
The friction between the soil and the wall is considered by means of the
friction angle , which is the angle to the normal of the total active thrust P a.
Depending on the relative movement between the soil and the wall, the force
Pa will be below or above the normal.
When the wall moves away from the soil and tilts (fig. 2.25), P a is above the
normal, with a component along the wall Pav which is a friction force and a
component normal to the wall Pah. In this case, which is the most common
one, can be taken as positive.
51
Fig. 2.25
There is also a case when the wall is founded on a very compressible soil
(fig. 2.26). As a result, the wall will settle. The friction force P av is opposing to
the downward movement of the wall. As a consequence, the force P a will be
below the normal.
Fig. 2.26
Computations performed with different values of have shown that has a
small influence on the magnitude of P a. Instead, the variation of the horizontal
component of Pa which is Pah = Pa cos can have a significant influence on
the stability conditions of the wall.
In general:
(2.34)
In practice, it is recommended:
< <
(2.35)
52
Fig. 2.27
A vertical wall limited by horizontal soil surface is considered (fig. 2.27). As
shown in the p. 1.2.1, the active state can be developed only within a wedge
of soil between the wall and a failure plane passing through the lower end of
the wall and at an angle of ( 45 o
Fig. 2.28
Fig. 2.29
53
If the wall deforms by rotation about its upper end (fig. 2.29) the condition for
the development of the active state is not satisfied since adequate strain in
the soil near the surface cannot take place. The soil adjacent to the wall in its
upper part cannot reach the limiting equilibrium and remains in elastic state.
The prism of soil limited by the failure surface moves downward. To render
possible such a movement, the failure surface should intercept the horizontal
ground surface under a right angle. The failure surface cannot be taken any
longer as a plane one, like in Coulombs theory, but with sufficient
approximation can be taken as a logarithmic spiral. The soil mass at the lower
end of the wall is obstructed to slip by the friction forces developed at the
contact with the part of the soil mass remained in elastic state. On that basis,
the soil mass which tends to slip transfers the pressure to the upper part of
the wall. This is an expression of the arching effect (the part of the soil mass
which can move is discharged by transferring the pressure to the upper part
which is blocked). Due to the arching effect, the distribution of the earth
pressure is no longer linear, but parabolic. The point of application of the total
active thrust is situated between 0,45 H and 0,55 H from the bottom of the
wall (fig. 2.30).
Fig. 2.30
2.3 Lateral earth pressure in the hypothesis of curved failure surfaces
2.3.1 Active case
A vertical wall limited by a horizontal ground surface is considered (fig. 2.31).
The crest of the wall can move upward and away from the soil, thus creating
the condition for the development of the active state. Due to the friction
between the soil and the wall, shear stresses occur along the wall.
Consequently, the vertical face of the wall cannot be a principal direction, as
was the case in the Rankines theory. The last failure plane with an angle (
45
passing through the crest of the wall. Between this plane and the wall the
failure surfaces are curved, as a result of the influence of the shear stresses
54
along the wall. The failure surface is a mixed one, composed by a curve and
Fig. 2.31
In the active case, the curvature is slight and the error involved in assuming a
plane surface is relatively small, less than 5%, which is fully allowable.
In conclusion, in the active case it is not justified to compute the lateral earth
pressure considering curved failure surfaces.
2.3.2 Passive case
A vertical wall limited by a horizontal ground surface is considered (fig. 2.32).
The crest of the wall can move downward and against the soil, thus creating
the condition for the development of the passive state.
The last plane corresponding to the Rankine passive zone passes through
the crest of wall. The wedge of soil influenced by the shear stresses on the
wall is much larger than in the active case. Between the results of the
computations performed using Coulombs theory (plane failure surface) and
those obtained in the Theory of Plasticity with curved failure surface,
differences are great not only concerning the shape of the failure surface but,
more important, concerning the magnitude of the total passive resistance
(40% or even more).
Fig. 2.32
55
Fig. 2.33
If the soil behind the wall has both internal friction and cohesion, the
computation is made in two steps.
a. soil with internal friction and weight ( 0,
0, c 0)
The failure plane drawn from the crest of the wall A, at an angle ( 45 ) to
the horizontal separates the Rankine passive zone from the zone of curved
failure surface. An arbitrary point D is selected on that plane (fig. 2.34).
Fig. 2.34
In order to find the centre 0 of the circular arc BD, a normal in the point D to
the line DC is drawn as well as a normal in the middle of the chord BD. At the
intersection of the two normals, the centre 0 is found. When the wall
displacement is such that the total passive resistance is fully mobilized, the
soil within the triangle ADC is in the passive Rankine state, both angles EAD
56
and ECD being ( 45 2 ). The horizontal force (E p) on the vertical plane DE,
therefore, is the Rankine passive value given by equation (2.36), acting
horizontally at a distance DE/3 above the point D.
Ep =
Ep
2
1
DE tan 2 ( 45 o )
2
2
(2.36)
It is necessary, then, to analyse only the stability of the soil mass ABDE. The
forces to be considered are:
the weight (G) of ABDE, acting through the centroid of the section;
the force Ep on DE;
the reaction (P) to the face between the soil and the wall, acting at angle
above the normal and at a distance of AB/3 above B;
the reaction (Q) on the failure surface BD when the shear strength along
the circular arc BD is fully mobilized, the reaction Q being assumed to act
at an angle to the normal.
In order to find the direction and the point of application of the reaction Q, the
arc BD is divided in a serie of elements of length ds, with the elementary
reaction qds applied making the angle with the normal (the radius) in the
point of application.
From the centre 0 of the circular arc of radius R, a normal OT is drawn on the
line AD, which is the line of action of the elementary reaction qds in the point
D.
OT = OD sin = R sin
A circle of radius r = R sin is drawn. This circle is referred to as the friction
circle or -circle.
Since all elementary reactions make the angle with the radius R, it is
obvious that they are tangent to the - circle, which can be defined as the
locus of the points of tangency of the lines of action of the elementary
reactions qds.
In order to find the direction of the force Q, which is the resultant of the
elementary reactions qds, the moment in respect to the centre 0 is taken:
Q d qds R sin R sin qds
d R sin
qds
Q
(2.37)
qds Q , since a polygonal contour is larger than the straight line between
the extremities of the contour (fig. 2.35).
57
Fig. 2.35
It results than d > R sin
qds
1 and d = R sin (the error involved is small and
By approximation,
Q
Fig. 2.36
The point D on the failure line passing through A was chosen arbitrarily. The
analysis must be repeated for a number of failure surfaces corresponding to
other points D1, D2etc. By plotting the values of P against the ground
surface taken as a reference line (fig. 2.37) a curve is obtained. The
tangential to the curves gives Pmin = Pp.
58
Fig. 2.37
b. soil having only cohesion
The superposition of effects will be used:
Ptot Pp Pp'
p
(2.38)
where Pp is the passive resistance due to the internal friction and to the
weight of the soil (determined as it was shown)
Pp' is the additional passive resistance due to cohesion.
'
In order to find Pp for the same failure surface BDC, the following procedure
is used (fig. 2.38).
Fig. 2.38
The passive resistance corresponding to the wedge EDC is:
59
E p' 2 c DE tan ( 45 o
(2.39)
The forces to be considered to analyse the stability of the soil mass ABDE
are:
Ca - the resultant of the adhesion between the soil and the wall; Ca = ca
AB
Fig. 2.39
Therefore, the force C = c
To find the line of action of the force C, the moment in respect to the centre of
the circular arc is considered.
BD
C a
cds R c R ds c R BD
c BD a c R BD
60
aR
BD
BD
(2.40)
Where: a is the distance from the line of action of the force C to the centre 0.
The force P ' is obtained as follows:
- Forces Ca and C are composed giving a resultant Ct
- Ct is composed with E 'p giving a resultant S '
- The forces S ' , P ' and Q ' must intersect, therefore the line of action of
Q ' must pass through the intersection of S ' and P ' and be tangential
to the -circle. The polygon of forces is completed and P ' is
determined.
The analysis is repeated for the same points D 1, D2, for which P was
previously determined (for the case 0, 0, c 0) .
By summing P and P ' for each trial and representing the forces P+ P '
against the reference line (fig. 2.40) a curve of variation is drawn. The
tangential to the curve parallel to the reference line defines Ptot min = Ptot p,
which is the total passive resistance including the effect of both internal
friction and cohesion.
Fig. 2.40
61