EM 1110-2-2504

31 Mar 94

Chapter 5 anchored wall, the anchor is assumed to prevent any

System Stability lateral motion at the anchor elevation. Other assump-
tions are discussed in the following paragraphs.
5-1. Modes of Failure
b. Preliminary data. The following preliminary
information must be established before design of the
The loads exerted on wall/soil system tend to produce a
system can commence.
variety of potential failure modes. These failure modes,
the evaluation of the loads on the system, and selection
(1) Elevation at the top of the sheet piling.
of certain system parameters to prevent failure are dis-
cussed in this chapter.
(2) The ground surface profile extending to a mini-
mum distance of 10 times the exposed height of the
a. Deep-seated failure. A potential rotational fail-
wall on either side.
ure of an entire soil mass containing an anchored or
cantilever wall is illustrated in Figure 5-1. This poten-
(3) The soil profile on each side of the wall includ-
tial failure is independent of the structural characteristics
ing location and slope of subsurface layer boundaries,
of the wall and/or anchor. The adequacy of the system
strength parameters (angle of internal friction φ,
(i.e. factor of safety) against this mode of failure should
cohesive strength c, angle of wall friction δ, and
be assessed by the geotechnical engineer through con-
wall/soil adhesion) and unit weight for each layer to a
vential analyses for slope stability (EM 1110-2-1902).
depth below the dredge line not less than five times the
This type of failure cannot be remedied by increasing
exposed height of the wall on each side.
the depth of penetration nor by repositioning the anchor.
The only recourse when this type of failure is antici-
(4) Water elevation on each side of the wall and
pated is to change the geometry of retained material or
seepage characteristics.
improve the soil strengths.
(5) Magnitudes and locations of surface surcharge
b. Rotational failure due to inadequate pile pene-
loads.
tration. Lateral soil and/or water pressures exerted on
the wall tend to cause rigid body rotation of a cantilever
(6) Magnitudes and locations of external loads
or anchored wall as illustrated in Figure 5-2. This type
applied directly to the wall.
of failure is prevented by adequate penetration of the
piling in a cantilever wall or by a proper combination of
c. Load cases. The loads applied to a wall fluctuate
penetration and anchor position for an anchored wall.
during its service life. Consequently, several loading
conditions must be defined within the context of the
c. Other failure modes. Failure of the system may
primary function of the wall. As a minimum, a cooper-
be initiated by overstressing of the sheet piling and/or
ative effort among structural, geotechnical, and hydrau-
anchor components as illustrated in Figures 5-3 and 5-4.
lic engineers should identify the load cases outlined to
Design of the anchorage to preclude the failure depicted
be considered in the design.
in Figure 5-4a is discussed later in this chapter. Design
of the structural components of the system is discussed
(1) Usual conditions. The loads associated with this
in Chapter 6.
condition are those most frequently experienced by the
system in performing its primary function throughout its
5-2. Design for Rotational Stability
service life. The loads may be of a long-term sustained
nature or of an intermittent, but repetitive, nature. The
a. Assumptions. Rotational stability of a cantilever
fundamental design of the system should be optimized
wall is governed by the depth of penetration of the
for these loads. Conservative factors of safety should
piling or by a combination of penetration and anchor
be employed for this condition.
position for an anchored wall. Because of the complex-
ity of behavior of the wall/soil system, a number of
(2) Unusual conditions. Construction and/or main-
simplifying assumptions are employed in the classical
tenance operations may produce loads of infrequent
design techniques. Foremost of these assumptions is
occurrence and are short duration which exceed those of
that the deformations of the system are sufficient to
the usual condition. Wherever possible, the sequence of
produce limiting active and passive earth pressures at
operations should be specified to limit the magnitudes
any point on the wall/soil interface. In the design of the

5-1
EM 1110-2-2504
31 Mar 94

Figure 5-1. Deep-seated failure

Figure 5-2. Rotational failure due to inadequate penetration

5-2
 EM 1110-2-2504
31 Mar 94

Figure 5-3. Flexural failure of sheet piling

and duration of loading, and the performance of the wall the experience of the designer and the reliability of the
should be carefully monitored to prevent permanent values assigned to the various system parameters. A
damage. Lower factors of safety or higher material procedure which has gained acceptance in the Corps of
stresses may be used for these conditions with the intent Engineers is to apply a factor of safety (strength reduc-
that the system should experience no more than tion factor) to the soil strength parameters φ and c while
cosmetic damage. using "best estimates" for other quantities. Because
passive pressures calculated by the procedures described
(3) Extreme conditions. A worst-case scenario in Chapter 4 are less likely to be fully developed than
representing the widest deviation from the usual loading active pressures on the retaining side, the current
condition should be used to assess the loads for this practice is to evaluate passive pressures using "effec-
case. The design should allow the system to sustain tive" values of φ and c given by
these loads without experiencing catastrophic collapse
but with the acceptance of possible major damage which
tan(φeff) tan(φ) / FSP (5-1)
requires rehabilitation or replacement. To contrast usual
and extreme conditions, the effects of a hurricane on a
hurricane protection wall would be the "usual" condition
governing the design, while the loads of the same hurri- and
cane on an embankment retaining wall would be
ceff c / FSP (5-2)
"extreme."

d. Factors of safety for stability. A variety of

methods for introducing "factors of safety" into the where
design process have been proposed; however, no
universal procedure has emerged. In general, the design FSP = factor of safety for passive pressures
should contain a degree of conservatism consistent with

5-3
EM 1110-2-2504
31 Mar 94

Figure 5-4. Anchorage failures

5-4
 EM 1110-2-2504
31 Mar 94

Minimum recommended values of FSP are given in For convenience in calculations for stability, the
Table 5-1. A factor of safety FSA may be applied for individual distributions are combined into "net" pressure
active pressures, however it is considered sufficient to distributions according to:
use an FSA = 1 in most cases unless deformations of
the wall are restricted. "NET ACTIVE" PRESSURE = retained side active
soil pressure
- dredge side passive soil
Table 5-1
Minimum Safety Factors for Determining the Depth pressure
of Penetration Applied to the Passive Pressures + net water pressure
(+ pressure due to
Loading Case Fine-Grain Soils Free-Draining Soils
retained side surcharge)
Floodwalls (- pressure due to dredge
side surcharge)
Usual 1.50 Q-Case 1.50 S-Case
1.10 S-Case "NET PASSIVE" PRESSURE = retained side passive
Unusual 1.25 Q-Case 1.25 S-Case
soil pressure
1.10 S-Case - dredge side active soil
pressure
Extreme 1.10 Q-Case 1.10 S-Case + net water pressure
1.10 S-Case (+ pressure due to
retained side surcharge)
Retaining Walls
(- pressure due to dredge
Usual 2.00 Q-Case 1.50 S-Case side surcharge)
1.50 S-Case
In these definitions of net pressure distributions, positive
Unusual 1.75 Q-Case 1.25 S-Case pressures tend to move the wall toward the dredge side.
1.25 S-Case
Typical net pressure diagrams are illustrated in
Extreme 1.50 Q-Case 1.10 S-Case Figure 5-5.
1.10 S-Case
f. Stability design for cantilever walls. It is assumed
that a cantilever wall rotates as a rigid body about some
e. Net pressure distributions. Evaluations of the point in its embedded length as illustrated in Fig-
pressures by the processes described in Chapter 4 result ure 5-2a. This assumption implies that the wall is
in a number of pressure distributions. subjected to the net active pressure distribution from the
top of the wall down to a point (subsequently called the
(1) Active soil pressures due to retained side soil. "transition point") near the point of zero displacement.
The design pressure distribution is then assumed to vary
(2) Passive soil pressures due to retained side soil. linearly from the net active pressure at the transition
point to the full net passive pressure at the bottom of
(3) Pressures due to surcharge loads on retained the wall. The design pressure distribution is illustrated
side surface. (Effects of surcharge loads are included in in Figure 5-6. Equilibrium of the wall requires that the
the soil pressures when a wedge method is used.) sum of horizontal forces and the sum of moments about
any point must both be equal to zero. The two
(4) Active soil pressures due to dredge side soil. equilibrium equations may be solved for the location of
the transition point (i.e. the distance z in Figure 5-6) and
(5) Passive soil pressures due to dredge side soil. the required depth of penetration (distance d in Fig-
ure 5-6). Because the simultaneous equations are non-
(6) Pressures due to surcharge loads on dredge side linear in z and d, a trial and error solution is required.
surface.
g. Stability design for anchored walls. Several
(7) Net water pressures due to differential head. methods for anchored wall design have been proposed

5-5
EM 1110-2-2504
31 Mar 94

Figure 5-5. Typical net pressure distributions

and classified as the "Free Earth" method (implied in penetration and anchor force, it will be necessary to
Figure 5-2b) and variations of the "Fixed Earth" hypoth- consider several anchor positions to arrive at the optimal
esis. Research and experience over the years have combination. For an initial estimate, the anchor may be
shown that walls designed by the Free Earth method are assumed to lie at a distance below the top of the wall
sufficently stable walls with less penetration than those equal to one-fourth to one-third of the exposed wall
designed by the Fixed Earth method. Because of the height.
flexibility of the sheet piling, the Free Earth method
predicts larger moments than those that actually occur. h. Anchor design. The anchor force calculated in
This shortcoming of the Free Earth method is overcome the stability analysis was obtained from equilibrium of a
by using Rowe’s moment reduction curves, as described typical 1-foot slice of the wall. In the actual system the
in Chapter 6. In the Free Earth method, the anchor is anchor support is provided by discrete tie rods attached
assumed to be a rigid simple support about which the to the wall through wales and to another support
wall rotates as a rigid body as shown in Figure 5-2b. mechanism (termed the "anchor" herein) at their ends
Despite the tendency of the wall to produce a passive and remote from the wall. Structural design of the tie
condition in the retained soil above the anchor, it is rods and wales is discussed in Chapter 6. A variety of
assumed that the wall is only subjected to the net active anchor configurations are illustrated in Figure 2-2.
pressure distribution as illustrated in Figure 5-7. The Capacities of some anchor configurations are discussed
required depth of penetration (d in Figure 5-7) is deter- in the following paragraphs. The soil strength
mined from the equilibrium requirement that the sum of parameters appearing in the equations associated with
moments about the anchor must be zero. After the anchor design should be consistent with the properties
depth of penetration has been determined, the anchor (S-case or Q-case) used for stability design. In all cases
force is obtained from equilibrium of horizontal forces. the capacity of the anchor should be sufficient to
Because the position of the anchor affects both depth of develop the yield strength of the tie rods (Chapter 6).

5-6
 EM 1110-2-2504
31 Mar 94

Figure 5-6. Design pressure distribution for cantilever wall

(1) Continuous anchors. A continuous anchor con- Figure 5-8) must be available. In addition, if the anchor
sists of a sheet pile or concrete wall installed parallel to wall intersects the line ac in Figure 5-8, interaction
the retaining wall as illustrated in Figures 2-2a and 2-2b. between the anchor wall and the retaining wall may
The continuous anchor derives its resistance from differ- increase the soil pressures on the retaining wall, thus
ential passive and active pressures produced by interac- invalidating the previous stability analysis. For non-
tion with the surrounding soil. homogeneous soil systems, the boundaries defining
minimum spacing of the anchor wall may be estimated
(a) Anchor location. The minimum distance from by the procedures used in the "Fixed Surface" wedge
the retaining wall at which an anchor wall must be method described in CWALSHT User’s Guide
placed to develop its full capacity is illustrated in Fig- (USAEWES 1990).
ure 5-8 for a homogeneous soil system. Under the
assumptions employed in the stability analysis of the (b) Full anchor capacity. Active and passive pres-
retaining wall, a zone of soil (bounded by line ab in sures developed on the anchor wall are shown in Fig-
Figure 5-8) behind the retaining wall is at its limiting ure 5-9 for a homogeneous soil system where h/H is 1/3
active state. To permit development of passive pres- to 1/2 (Teng (1962) and Terzaghi (1943)). The capacity
sures, an additional zone of soil (bounded by line bc in of the anchor wall is given by

5-7
EM 1110-2-2504
31 Mar 94

Figure 5-7. Design pressure distribution for free earth design of anchored walls

Ca PP PA (5-3) PP γH 2 KP /2 (5-4)

where and

Ca = anchor wall capacity per foot of anchor wall PA γH 2 KA /2 (5-5)

PP = resultant of the passive pressures in front of the

anchor wall where KP and KA are passive and active earth pressure
coefficients given in Equations 4-3 and 4-4 evaluated
PA = resultant of the active pressures in back of the with the same effective angle of internal friction used
anchor wall for stability analysis of the retaining wall but with zero
wall friction. For homogeneous soils with Q-case
For homogeneous soils with S-case strengths strength parameters, (KA = KP = 1)

5-8
 EM 1110-2-2504
31 Mar 94

Figure 5-8. Minimum anchor - wall spacing for full passive anchor resistance in homogeneous soil

Figure 5-9. Resistance of continuous anchor wall

5-9
EM 1110-2-2504
31 Mar 94

where
γH 2 (5-6)
PP 2cH
2 L= length of the dead man parallel to the
retaining wall

and B= thickness of the deadman perpendicular

to the retaining wall

γH 2 2c 2 (5-7) PA and PP = resultants of active and passive soil

PA 2cH
2 γ pressures (Equations 5-4 through 5-7),
respectively

where c is the effective soil cohesive strength used for φ and c = effective (factored) angle of internal
stability analysis of the retaining wall. friction and cohesive strength,
respectively
(c) Reduced anchor wall capacity. When physical
constraints require violation of the minimum spacing KP and KA = passive and active earth pressure coeffi-
between anchor wall and retaining wall, the attendant cients evaluated for effective strengths
reduced anchor wall capacity should be evaluated by the (Equations 4.3 and 4.4)
procedures discussed by Terzaghi (1934).
Ko = at-rest pressure coefficient which may
(d) Structural design of sheet pile and concrete be taken as
anchor walls. Sheet pile anchor walls should be
designed for maximum bending moment and shear
Ko 1 sin(φ) (5-10)
under the stress limitations delineated in Chapter 6.
Concrete anchors should be designed under the Ameri-
can Concrete Institute (ACI) 318 (1983) specifications
for concrete structure in contact with the earth. (3) Anchors at large depth. Capacities of anchors at
large depth below the ground surface may be taken as
(2) Discontinuous anchors. Discontinuous anchors the bearing capacity of a footing located at a depth
(or dead men) are usually composed of relatively short equal to the midheight of the anchor (Terzaghi 1943).
walls or blocks of concrete. The stress distribution
ahead of a dead man is illustrated in Figure 5-10a and a (4) Grouted anchorage. Grouted anchorage consists
free-body diagram is shown in Figure 5-10b. The of tie rods or tendons installed in cased, drilled holes
capacity of a dead man near the ground surface for with their remote ends grouted into competent soil or
S-case strengths (c = 0) may be taken as rock as illustrated in Figures 2-2c and 5. The grouted
length must be fully outside the active wall zone (line
ab in Figure 5-5). Tie rods must be designed to resist
Ca L(PP PA ) (1/3) Ko γ KP the anchor force determined from wall stability analysis
(5-8) plus any preload applied for alignment or limitation of
3 initial deflections. The capacity of all grouted anchors,
KA H tan (φ) W tan (φ)
which should develop the yield strength of the tie rod,
must be verified by proof tests by loading to
110 percent of their required resistance. At least two
anchors should be subjected to performance tests by
and for Q-case strengths, (φ = 0) loading to 150 percent of their design capacity.

(5) Pile anchors. Capacities of anchors composed of

Ca L( PP PA ) 2cH 2 LBc (5-9)
tension piles or pile groups, Figure 2-2, should be evalu-
ated by the procedures set forth in EM 1110-2-2906.

5-10
 EM 1110-2-2504
31 Mar 94

Figure 5-10. Resistance of discontinuous anchor (dead man)

5-11
EM 1110-2-2504
31 Mar 94

Figure 5-11. Grouted anchors

5-12