SEGMENTAL BLOCK RETAINING WALLS
WITH COMBINATION GEOGRID AND
ANCHOR REINFORCEMENTS
Prepared by:
C.R. Lawson & T.W. Yee
TC Nicolon Asia, Kuala Lumpur, Malaysia
Chris_Lawson@tcn-asia.com
J-C. Choi
Samsung Industry Co., Ltd, Seoul, South Korea
chjongch@yahoo.co.kr
May 18, 2010
1
ABSTRACT: Reinforced segmental block retaining walls can be constructed where the
reinforced fill zone is constrained by using a combination of geogrid reinforcement and anchor
reinforcement. The paper presents the design theory for reinforced soil walls where the
reinforced fill zone is constrained and where anchors are used to dissipate the reinforcement
residual tensile stresses beyond the reinforced fill zone. A case study is also presented where
this technique was successfully used.
INTRODUCTION
Geogrid reinforced segmental block retaining walls have become an accepted practice for
economical retaining wall construction. For this retaining wall system the segmental blocks
act as the wall facing, and provide some stability, with the majority of the stability provided by
the geogrid reinforced fill behind the segmental block facing. For stability, the geogrid
reinforcement has to have adequate design strength, be located at specific vertical spacings
in the reinforced fill, and extend an adequate distance into the reinforced fill. While geogrid
reinforcement design strengths and vertical spacings are easy to attain, the required
reinforcement lengths into the reinforced fill may be difficult to attain if there are outcrops of
heavily overconsolidated soils, soft rocks or hard rocks in close vicinity to the retaining wall.
In these instances it can be impractical to excavate the heavily overconsolidated soil or rock
and thus, some additional means must be found to fully dissipate the tensile stresses induced
in the reinforcement over its truncated length.
Figure 1. Geogrid reinforced segmental block wall with a constrained reinforced fill zone
One technique is to anchor the geogrid reinforcement into the heavily overconsolidated soil
or rock zone at the rear of the reinforced fill zone, Figure 1. The anchors act to fully dissipate
2
any residual tensile stresses at the rear of the geogrid reinforcement into the heavily
overconsolidated soil or rock zone. A comparison between a conventional geogrid reinforced
segmental block wall and one with a combination of geogrid and anchor reinforcements is
shown in Figure 2. For the conventional reinforced segmental block wall the extent of the
reinforced fill zone is adequate to account for local and external stability requirements.
However, the segmental wall with the constrained reinforced fill zone cannot generate
adequate internal stability in terms of geogrid reinforcement bond because of the truncated
reinforcement lengths. By attaching anchors to the geogrid reinforcements and inserting them
into the heavily overconsolidated soil or rock zone residual tensile stresses in the geogrid
reinforcement can be fully dissipated, and the retaining wall can achieve the required stability.
Figure 2. Conventional geogrid reinforced segmental block wall and one with a combination
of geogrid and anchor reinforcements
The general representation of the constrained reinforced soil retaining wall problem is shown
in Figure 3. Layers of geogrid reinforcement lying within the reinforced fill zone transfer tensile
stresses from the vicinity of the wall face, of height H, and dissipate these within the reinforced
fill, of base width Lb. The extent of the reinforced fill is constrained by a rigid zone, inclined at
an angle α to the horizontal, at the rear of the reinforced fill zone. If the tensile stresses in the
3
geogrid reinforcements cannot dissipate before reaching the rigid zone then anchors
attached to the geogrid reinforcements are inserted into the rigid zone to provide the required
residual tensile stress dissipation. Depending on the geometry, anchors may, or may not, be
required for the full height of the wall.
EARTH PRESSURES ACTING ON THE REAR OF THE SEGMENTAL BLOCK WALL
FACING
The vertical stresses acting within the reinforced fill are due to the self weight of the reinforced
fill plus surcharge, plus other additional vertical stresses due to external vertical loads acting
on the reinforced fill. Since the rigid zone at the rear of the reinforced fill zone does not impart
stresses onto the reinforced fill (unlike retained soil) there is no influence from this zone on
the vertical stresses in the reinforced fill zone. Further, while the interface of the rigid zone
may be considered an unyielding surface, the interface at the wall face is a yielding surface
due to the reinforcement connections with the wall face generally being made with extensible
geogrids. Consequently, the presence of this yielding surface at the wall face prevents any
arching occurring within the constrained zone of the reinforced fill. Thus, the vertical stresses
acting within the reinforced fill zone can be considered uniform across the full width of the
reinforced fill, and are due to the full self weight of the reinforced fill plus any surcharge and
external vertical load effects.
Figure 3. General representation of the reinforced soil retaining wall problem with
constrained reinforced fill zone
4
Figure 4. Maximum vertical stress at the base of the constrained reinforced soil wall
The maximum vertical stress acting across the base of the reinforced soil wall is shown in
Figure 4. For the reasons state above, the maximum vertical stress is uniform across the
width of the reinforced fill zone and is equal to the full self weight of the reinforced fill plus
any surcharge or other vertical external loading effects.
In determining the gross horizontal equilibrium force acting on the wall face the constrained
nature of the reinforced fill must be considered. Within the constrained reinforced fill zone the
active failure wedge may not be able to develop fully because of the relative close proximity
of the rigid zone at the rear of the reinforced fill. The general geometry of the problem is
shown in Figure 5a where the extent of the reinforced fill zone is described in terms of H, the
wall height, and Lt, the width of the top of the fill zone. While the boundary of the rigid zone
is shown as vertical in Figure 5a, in fact, the following analysis is also applicable where the
rigid zone boundary is inclined (i.e, at angle α in Figure 4).
The forces acting on the fill zone are shown in Figure 5b. The destabilizing force is due to the
weight of the fill, W, within the potential failure surface. For simplicity, it is assumed that the
equilibrium force acting on the rear of the wall face is horizontal and the wall face is vertical.
Thus, the equations for equilibrium are:
If Lt /H cot 45 o φ /2 ;
Ph Lt L sinθ tanφ cosθ
K 2 t tanθ (1a)
cosθ tanφ sinθ
2
0.5γ. H H
If Lt /H cot 45 o φ /2
Ph
0.5γ. 2
K cot 2 45 o φ /2 K a (1b)
5
where, the above variables are described in Figure 5b, and K is the horizontal force coefficient
acting on the rear of the wall facing.
Figure 5c shows values of K for various Lt/H ratios for a fill internal friction angle φ’ = 30°. For
Lt/H ratios greater than 0.5 the full active wedge can develop wholly within the fill zone and
2
hence the value of K= Ka= cot (45° + φ’/2), Equation 1b. For Lt/H ratios less than 0.5 the full
active wedge cannot develop wholly within the fill zone and hence the value of K decreases
for decreasing ratios of Lt/H. The value of K for a constant wedge angle θ = 45°+φ’/2 is shown
plotted in Figure 5c for various values of Lt/H. However, this value of K is not the maximum
value that can be attained for a specific Lt/H ratio. The reason for this is that the constrained
fill forces the critical failure surface to a greater wedge angle θ than what would normally be
the case under full active conditions. This results in a higher K value than when the wedge
angle θ = 45°+φ’/2 is used. This maximum horizontal force coefficient Kmax is shown plotted
in Figure 5c for fill internal friction angle φ’= 30°. There is a significant difference between this
Kmax value and the K value determined from a constant wedge angle θ = 45°+φ’/2 value
especially at low values of Lt/H.
Figure 5d shows the plot of the wedge angle θ that yields the maximum horizontal force
coefficient Kmax for a fill with internal friction angle φ’ = 30°. For Lt/H ratios greater than 0.5
where the full active wedge can develop within the fill the wedge angle is θ = 45°+φ’/2. For
Lt/H ratios less than 0.5 the wedge angle θ coincides closely, but not exactly, with the juncture
of the fill and the rigid zone at the top of the wall.
Figure 5e shows the calculated maximum horizontal force coefficients Kmax for fill types in the
range φ’= 25° to 45°. Figure 5f shows the wedge angle θ that may be used in Equation 1a to
calculate the maximum horizontal force coefficients Kmax for fill types in the range φ’ = 25° to
45°.
Differentiating the gross horizontal force Ph with respect to depth at the wall face yields the
horizontal stress distribution acting on the rear of the wall face in Figure 6.
For 0 z Lt tan 45 o φ /2
σ h cot 2 45 / 2γz w s (2a)
For Lt tan 45 o φ /2 z H
Lt L sinθ tanφ cosθ
σh 2 t tanθ γz w s (2b)
z z cosθ tanφ sinθ
6
Figure 5. Horizontal force coefficient acting on rear of wall facing due to weight of reinforced
fill
7
Figure 6. Horizontal stress distribution acting on rear of wall facing due to self weight and
surcharge and the gross horizontal force Ph is;
Lt L sinθ tanφ cosθ
Ph 0.5 2 t tanθ z γz 2w s (2c)
z z cosθ tanφ sinθ
where, θ is the appropriate critical wedge angle obtained from Figure 5f.
In the region between σho and σh1 in Figure 6 the horizontal stress distribution increases
linearly with depth while in the region between σh1 and σh2 the horizontal stress distribution
increases logarithmically with depth.
TENSIONS IN THE REINFORCEMENTS AND ANCHORS
Tensions are generated in the reinforcements and anchors which provide internal stability for
the reinforced soil wall. Figure 7a shows the stress regime at level j in the wall. A vertical
stress σvj is generated due to the weight of the fill plus surcharge and other external loads.
This vertical stress imparts a horizontal stress on the rear of the wall facing which is resisted
by the generation of a tension force in the reinforcement, Tmaxj. If this tension force is not
dissipated on reaching the end of the reinforcement at the boundary with the rigid zone then
the residual tension is transferred to the anchor, Panj, for dissipation within the rigid zone.
The location of maximum tension within each reinforcement layer in the reinforced soil wall
coincides with the juncture between the active and passive zones in the reinforced fill zone
defined by the wedge angle θ, Figure 7b.
The maximum tension generated in the reinforcement, Tmaxj, at level j in the wall is;
8
Tmaxj hj Svj (3)
where, σhj is the horizontal stress on the rear of the face of the wall at level j, and Svj is the
vertical spacing between the reinforcement layers at level j in the wall.
Figure 8 sets out the assumed tension distribution along the length of the geogrid
reinforcement at level j in the wall. For conservatism and ease of design it has been assumed
that the tension developed in the reinforcement at the wall face is equal to the maximum
tension generated, i.e. Tfacej = Tmaxj.
Figure 7. Internal stresses and location of maximum reinforcement tension
9
Figure 8. Distribution of tension along a layer of reinforcement
If the line of maximum tension occurs at some distance from the rigid zone boundary, Figure
8, then the tension generated in the reinforcement can dissipate within the passive zone of
the reinforced fill. The rate of tension dissipation within the passive zone is shown in Figure
9. It is governed by the overburden stress γh+ws, and the reinforcement/fill bond a’tanφ’.
The tension retained in the reinforcement at the rear of the passive zone, Trearj, at level j in
the wall is;
Trearj Tmax j 2 h j w s a tan Lej (4)
where, Tmaxj is the maximum tension occurring in the reinforcement at the juncture of the
active and passive zones at level j in the wall, γ is the unit weight of the reinforced fill, hj is
the height of the wall above level j, ws is the surcharge on top of the wall, a’ is the
reinforcement/fill bond coefficient, φ’ is the angle of internal friction of the reinforced fill, and
Lej is the reinforcement length in the passive zone at level j in the wall.
10
Figure 9. Assumed linear profile of reinforcement tension dissipation in the passive zone of
the reinforced fill
If the tension in the reinforcement has not been fully dissipated on reaching the boundary of
the rigid zone then anchors will be required to dissipate this residual tension within the rigid
zone. The magnitude of the working load in each anchor, Panj, will be;
Panj Trearj Shj sec (5)
where, Panj is the working load in the anchor at level j in the wall, Trearj is the tension in the
reinforcement at the rear of the passive zone at level j in the wall, Shj is the horizontal spacing
between adjacent anchors at level j in the wall, and ω is the angle of inclination to the
horizontal of the anchor at level j in the wall.
TYPES OF ANCHORS USED
The anchors used to connect the reinforcements into the rigid zone are relatively low
capacity, non-stressed anchors or soil nails. These can be of the form of bonded passive
anchors and nails, rotational dead-man tieback anchors, or driven nails. Consideration needs
to be given to the required design life of the anchors because they must perform for the full
design life of the reinforced soil wall. In some cases this design life may be as long as 120
years.
The connection between the anchor and the geogrid reinforcement is normally effected by a
circular steel bar or pipe. This type of connection requires the use of flexible geogrids with
negligible bending resistance to ensure stress concentrations do not occur at the connection.
REINFORCED SEGMENTAL BLOCK RETAINING WALL AT PA JU, SOUTH KOREA
11
The landscaping and earthworks for an apartment complex at Pa Ju, North of Seoul, South
Korea required the construction of several reinforced segmental block retaining walls. At one
location a 130 m long retaining wall was required, ranging in height from 3 m to 8 m.
The area where this wall was to be constructed consisted of partially decomposed quartzite.
In order to minimize risk to an adjacent building the owner decided not to excavate to
construct a conventional geogrid reinforced segmental block wall. Instead a combination
geogrid and anchor reinforced segmental block wall was constructed.
12
13
Figure 10a shows the proposed retaining wall geometry in relation to the location of the
quartzite stratum whose surface was inclined approximately 60° to the horizontal. The design
called for a wall of exposed height 5.6 m with a crest width of 3.0 m.
Figure 10b shows the various design parameters used for the constrained reinforced soil
wall. The ratio Lt/H = 3.8/6.1 = 0.62 gives a critical wedge angle θ = 45°+φ’/2 = 62.5° (Figure
5f). Because of the wall geometry the full active wedge can develop within the reinforced fill
zone, and the line of maximum reinforcement tension is shown plotted in Figure 10b. The
passive zone within the reinforced fill is very limited and corresponds approximately with the
boundary of the rigid zone (the quartzite stratum). Consequently, for the design of the wall it
was assumed that the maximum tension developed in the reinforcement occurs at the
boundary of the rigid zone, i.e. at the face of the quartzite stratum.
The segmental wall facing consisted of Anchor Vertica Pro blocks. The horizontal stress
distribution acting against the rear of the wall facing was determined in accordance with
Equation 2, and the appropriate geogrid vertical spacing was determined on the basis of
geogrid design strength and the connection capacity between the geogrid reinforcement and
the segmental block facing. For this wall it was planned to economize on the number of
anchors by utilizing one row of anchors for every two layers of geogrid reinforcement, Figures
10c and 10d. The original design called for 260 Mantaray MR2 earth anchors to be installed
into the partially decomposed quartzite stratum. These were to be installed by pre-auguring
into the partially decomposed quartzite back slope prior to anchor insertion. The detailed
layout of the reinforced segmental block wall is shown in Figure 10c.
The connection between the anchors and the geogrid reinforcement was effected by the use
of a 75 mm diameter galvanized steel pipe, Figure 10d. Miragrid geogrids were used as the
geogrid reinforcement as these exhibited the required long-term design strengths and had
the required level of bending flexibility which enabled them to pass easily around the
galvanized steel pipe without attracting additional tensile stress.
During construction it was found that the quartzite stratum was harder than originally
anticipated. Consequently, drilled rock bolts had to be substituted for the majority of the
Mantaray earth anchors. A total of 208 rock bolts were used in addition to 52 Mantaray MR2
earth anchors.
CONCLUSIONS
The paper provides an analytical model for reinforced soil walls where the reinforcements
consist of a combination of geogrids and anchors (or nails). The technique is particularly
suited to situations where a rigid zone constrains the extent of the reinforced fill zone. In this
case residual tensions in the geogrid reinforcements within the reinforced fill zone are
transferred into the rigid zone by means of anchors or nails. A major consideration is that all
components must fulfill the design life requirements of the retaining wall.
The geogrid and anchor reinforced segmental block wall at Pa Ju, South Korea has
demonstrated that this combination reinforcement technique can be used successfully.
Disclaimer: TenCate assumes no liability for the accuracy or completeness of this
information or for the ultimate use by the purchaser. TenCate disclaims any and all express,
implied, or statutory standards, warranties or guarantees, including without limitation any
14
implied warranty as to merchantability or fitness for a particular purpose or arising from a
course of dealing or usage of trade as to any equipment, materials, or information furnished
herewith. This document should not be construed as engineering advice.
© 2010 TenCate Geosynthetics North America
15
