1803
Technical Committee 205 /
Comité technique 205
A fundamental difference between design of this foundation
solution and that of a piled mat is the presence of negative skin
friction, as the layer of gravel pushes down on the soil, which
compresses over time between the piles. The authors review
French and German prescriptions for design of this type of
foundation solution.
Shallow
Pile foundation
„
rigid inclusions
“
Combined pile‐raft
foundation
Load transfer layer
Figure 2 Comparison of common foundations solutions with pile-
supported blanket solution (after Katzenback et al. 2013).
5.3
Discussion
planes with equal
settlement
Load transfer layer
rigid or flexible raft foundation
(or embankement)
neutral
plane
Figure 3 Occurrence of negative skin friction along unreinforced
concrete piles caused by load applied by the layer of gravel on top of the
column on the soil between the columns (after Katzenback et al. 2013).
6 EARTH RETENTION
6.1
General Remarks
Fundamental work on the probabilistic analysis of retaining
walls has tended to focus on mechanically stabilized earth
(MSE) walls. This work has resulted in part from the interest
sparked by the AASHTO LRFD mandate in North America and
the fact that MSE walls are widely used in transportation
infrastructure.
While retaining wall limit states in general should be treated
similarly to the slope stability limit state (because the pressure
on the walls results from shear surfaces that develop in the
backfill and that depend very much on spatial variability of
shear strength in the backfill), the presence of an interface (the
back of the wall) again allows the effects of the spatial
variability of soil variables in the backfill to show as variability
in contact pressure on the wall. Likewise, reinforcing elements
in MSE walls will develop pullout resistance and exert
stabilizing action on the backfill through the shear stresses (unit
resistances) between them and the soil; variability of these unit
resistances are then used in reliability analyses. In the case of
MSE walls, instrumentation of reinforcements, including near
the wall facing, provide data that can be used for estimation of
both earth pressures and unit interface resistance of reinforcing
elements (see, for example, Kim and Salgado 2012a,b).
An interesting issue in connection with retaining walls is the
choice of limit states that must be checked in design. The
traditional, idealized retaining wall limit states are sliding,
overturning, bearing capacity failure and general instability. In
MSE walls, the additional limit states of pullout and reinforcing
element rupture must also be checked. However, as pointed out
by Loukidis and Salgado (2012), realistic limit states tend to be
a composition of these idealized limit states. Equally interesting,
there is a relationship between mobilization of shear strength
(and consequently pressures on the wall) and wall movement,
and this relationship has implications for the setting up of
design situations. Merrifield et al. (2013) discuss aspects of this
issue as well, as did Loukidis and Salgado (2012) and Simpson
and Driscoll (1998).
6.2
Papers
Ho et al. (2013) used the design approach proposed in CIRIA
Report C580 in the design of a permanent cantilevered, large-
diameter, bored-pile wall (Figure 4) for the support of sloping
ground bordering a new road, which now exists in front of the
wall. This is the first known Hong Kong project in which a
permanent retaining structure was designed using the C580
design approach. The wall, which is approximately 110m long,
is made of 33 bored piles with diameter equal to 3.0m. Figure 5
shows a cross section of the wall as well as the original and
post-construction ground profile.
Figure 4 Bored-pile wall in Hong Kong (after Ho et a. 2013).
FILL
CDG
HDG
Figure 5 Cross section of bored pile wall in Hong Kong (after Ho et al.
2013).
The ground profile at the site consists essentially of fill and
weathered granite. In the original ground profile, the fill had a
maximum thickness of 8m and contained loose to medium
dense, gravely silty sand or sandy clayey silt, with some rock
and concrete fragments as well as domestic waste. Completely
decomposed granite (CDG) and highly decomposed granite
(HDG) are present below the fill. The granites in the area are
commonly fine- to medium-grained and greyish pink to pinkish
grey in their fresh state. The thickness of the CDG ranges from
5 m to 10 m with SPT blow counts ranging from 15 to 100. The
HDG occurs with SPT blow coutns greater than 100. The
maximum depth to moderately to slightly decomposed granites
(M/SDG) ranges from 15m to 40m below original ground level.
The water level was assumed at the excavation level on the
excavation side of the wall and within the the CDG layer on the
