Civil Engineering Reference
In-Depth Information
2.61m
2
of cap associated with it, fromwhich the effective diameter of each cap element
may be calculated as 1.82 m. The radius of influence of each pile is
r
m
=
22
.
2 m (from
equation (4.39), taking
v
=
0
.
1). Equation (5.29) then gives
α
rp
=
0
.
70
.
The overall
stiffness of the foundation may then be calculated as
k
f
=
5720 kN/mm with 24%
of the load being carried by the pile cap (equations (5.27) and (5.28)).
Cooke
et al
. (1981) report an overall load on the foundation of 156MN, fromwhich
the estimated average settlement may be calculated as 156
27 mm. The field
measurements indicate a long-term settlement of about 25 mm, with the proportion
of load carried by the pile cap dropping from an initial value of about 50%, down
to a long-term value of 23%. These values compare well with the estimated values
obtained above.
Randolph (1983) commented that the piles in this foundation design were only con-
tributing about 5% of their possible stiffness as individual piles. Halving the number
of piles under the foundation would have reduced the pile group stiffness by only 13%,
giving an estimated average settlement of 29 mm for the building, with some 31% of
the load being taken by the pile cap.
Padfield and Sharrock (1983) discuss an alternative design of the foundation for
Stonebridge Park (see Figure 5.28) which uses 40 piles, situated near the centre of
the raft, at a spacing of 3.2 m (7.1 pile diameters). Horikoshi and Randolph (1999)
suggest even fewer piles, with a 18 piles on a 3
/
5
.
72
=
6 grid over the central 25% of the
raft. The proposed piles were 0.5 m in diameter, penetrating 28 m below the raft,
at a spacing of 3.5 to 4 m. Both of these schemes appear to eliminate virtually all
differential settlements, even for a flexible raft.
An estimate of the average settlement of the foundation proposed by Horikoshi
and Randolph (1999) is based on their pile group stiffness of
k
p
∼
×
3000 MN/m (esti-
mated by two independent calculations) and an interaction factor
α
rp
of 0.66. Using
the raft stiffness given above and the relationship in equation (5.27), the overall
stiffness of the piled raft is calculated as
k
pr
=
3710 MN/m. The average settle-
ment of the piled raft is therefore estimated as 42 mm under the design load of
156.6 MN.
This example serves to demonstrate that significant savings may be made in foun-
dation design by eliminating piles that are not needed from a stability point of view.
Relatively small increases in average settlement will occur due to the reduction in
the number of piles, but this is offset by a significant reduction in the differential
settlements.
There are of course many considerations that come into the design of a piled founda-
tion. Combined pile and raft foundations are only viable where the surface soil layers
are reasonably competent, and even then the number and spacing of piles may well
be determined by the spacing of the columns carrying the structural load. The use of
one pile beneath each column is a convenient way of minimizing bending moments
in the pile cap. However, the piles may still be designed to take only a proportion of
the column loads, the remaining load being shed into a pile cap cast directly on the
ground. In other situations, the principle of concentrating the pile support in the cen-
tral area of the raft may be followed by varying the pile length across the foundation.
A good example of that has been reported recently by Liew
et al
. (2002), where piles
of lengths varying from 36 m towards the centre down to 24 m at the edges were used
to support a 20 m diameter oil tank on soft ground.
