Deep Foundations
Deep Foundations
Piles used in marine structures are subjected to lateral loads from the impact of berthing ships
and from waves. Combinations of vertical and horizontal loads are carried where piles are
used to support retaining walls, bridge piers and abutments, and machinery foundations.
The basis of the ‘soil mechanics approach’ to calculating the carrying capacity of piles is that
the total resistance of the pile to compression loads is the sum of two components, namely
skin friction and end resistance. A pile in which the skin-frictional component predominates
is known as a friction pile, while a pile bearing on rock or some other hard incompressible
material is known as an end-bearing pile, as shown below.
Piles are often classified as “driven” or “bored”, but it is adequate in many situations, but
does not cope with many ‘hybrid” types in use. A better classification is “displacement” or
“non-displacement”.
Displacement piles: Generally driven. Soil is moved radially as pile shaft enters ground.
There may be a component of vertical movement.
Displacement piles in clay exhibit surface heave over a depth of up to 5 pile diameters, there
being little immediate volume of the clay. Excess pore pressure is generated next to the pile
1
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
so that the effective horizontal (radial) effective stress next to the pile is often very low,
hence reducing the energy to drive the pile. However, these excess pore pressures next to the
shaft can dissipate quickly (small drainage path) and hence pauses in driving. (e.g. overnight
or because of equipment malfunction) are accompanied by pore pressure dissipation and an
increase in pile resistance.
2
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
Displacement piles in sand tend to densify the sand, the extent depending on the initial in-situ
density.
Piles of relatively small cross-sectional area, such as “H-piles” are termed low-displacement
piles, and the effects of compaction (in sand) or heave (in clay) are reduced. This can be
useful if long piles are to be driven through granular deposits, or if the piles are at close
centres (where compaction would lead to driving difficulties), or where clay heave is a
problem.
Non-displacement piles: Generally bored. Lateral stresses in the ground are reduced during
excavation and only partly restored by concreting.
Non-displacement piles in clay suffer from reduced ground stresses and with no temporary
support (i.e. casing), the horizontal stress σr will drop to zero, with negative pore pressures
maintaining effective stress and the undrained condition. On dissipation, these suction will
cause reduction in σr’, and hence softening and swelling.
Non-displacement piles in sand require a casing, causing less effect on the in-situ stresses in
the soil surrounding the pile.
3
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
Some advantages and disadvantages associated with various pile types are given in the table
below:
4
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
The pile capacity is the sum of the end bearing capacity and shaft friction:
𝑄 = 𝑄𝑏 + 𝑄𝑠 … … . (2.1)
𝑙
πd2
= 𝑞𝑏 + 𝜋𝑑 ∫ 𝜏𝑠 𝑑𝑙 … . . (2.2)
4 0
πd2
= 𝑞 + 𝜋𝑑𝜏̅𝑠 𝑙
4 𝑏
The relative magnitudes of the shaft and base capacities will depend on the geometry of the
pile and the soil profile. Piles which penetrate a soft layer to found on a firmer stratum are
known as “end-bearing piles” and derive most of their capacity from their base Qb. Where no
particularly firm stratum is available to found piles on, the piles are known as “friction” or
“floating” piles.
In clays, the shaft capacity is generally the greater. In granular materials, or for underreamed
piles, the overall capacity is more evenly divided between the shaft and base.
Research has shown that the shaft capacity of a pile is mobilized at much smaller
displacement than the base capacity. This is the key to understanding the pile load-settlement
response and to interpreting pile test results.
5
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
Typically:
Consider the response below for a 0.6m diameter, 10m long bored pile in stiff clay:
Ultimate capacity=
Shaft Capacity=
Shaft load=
Base load=
and at working load, 95% of the working load is carried by the shaft. It is absolutely essential
to appreciate the mechanisms of load transfer shown in the figure above.
𝑄𝑏 + 𝑄𝑠
𝑄= … . . (2.3)
2.5
With a pile test:
𝑄𝑏 + 𝑄𝑠
𝑄= … . . (2.4)
2
For large diameter and underreamed piles without a pile test, use the lower of:
𝑄𝑏 + 𝑄𝑠 𝑄𝑏
𝑄= 𝑎𝑛𝑑 𝑄 = 𝑄𝑠 +
2.5 3
6
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
𝑞𝑏 = 𝑁𝑐 𝐶𝑢 … . (2.6)
With Nc=9. See Skempton’s chart in the Shallow Foundation Data Sheets: for a deep strip
footing, Nc=7.5. The shape factor for a circular or square footing is 1.2, giving an overall Nc
value of 1.2 x 7.5=9 for a pile of circular or square cross section.
A drained value of qb is not relevant. The undrained capacity is much less, and considerable
deformations would be required to mobilise the drained capacity anyway.
2.4.1 α-method
The shaft friction had been calculated historically using a total stress approach:
𝜏𝑠 = 𝛼𝐶𝑢 … … (2.7)
This is known as the α-method or total stress method for calculating shaft capacity and α is
known as the “shaft-adhesion factor”. The value of α is a function of the type of clay, the
installation method and the pile material and reduces with increasing value of undrained
shear strength cu. The value of α can be as much as 1.5 in soft sensitive clays and as low as
0.2 in very stiff clays (strength above 150kPa). The value of α also tends to be greater for
driven piles than for bored piles.
Because cu is not a soil constant, and depends on the method of measurement and due to the
uncertainties in the value of α, more recently an effective stress approach is being used.
Furthermore, since drainage path around the pile is very short, pore pressure next to the shaft
dissipate very quickly, so a drained analysis of shaft capacity is now considered by many to
be more appropriate. The approach is known as the β-method of the effective stress method.
7
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
2.4.2 β-method
The shaft friction is related to the horizontal effective stress by the equation:
The horizontal effective stress is related to the vertical effective stress by the equation:
Hence we write:
with
β = 𝐾 𝑡𝑎𝑛 𝛥 … … . (2.11)
Parameter K is a horizontal earth pressure coefficient, and is a function of the soil properties
and the installation method.
1 𝑙𝑜
𝜏̅𝑠 = ∫ 𝛽𝜎𝑣 ′ 𝑑𝑙 = 𝛽̅ ̅̅̅̅
𝜎𝑣 ′ … . (2.12)
𝑙 0
̅̅̅
𝜏̅𝑠 = 𝛼𝐶 𝑢 … … (2.13)
𝛽̅ can be determined from pile tests, as can 𝛼̅. However, clearly σv’ is known much more
precisely than ̅̅̅,
𝑐𝑢 so the effective stress method has a clear advantage over the total stress
method.
In general, the value of α or β will depend not only on the shear strength of the soil cu, but
also on its past history and overconsolidation ratio, but the details are outside the scope of
this course.
8
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
𝑞𝑏 = 𝑁𝑞 𝜎𝑣′ … . (2.14)
9
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
The value of K depends on Ko , the method of installation and initial density of the sand. In
coarse-grained materials (sands and gravels) the parameter K is usually expressed in terms of
K0, the coefficient of lateral earth pressure at rest. This parameter depends on the soil type
and stress history (quantified by φ′ and OCR). For cast-in-situ piles 0.7 < K/ K0 < 1.0, while
for displacement piles K/K0 may be as high as 2.0.
𝑁𝑞
K≈
50
≈1.5 at low stress levels
≈1 or lower at greater depths, with an average of 1.2 from the two values.
10
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
11
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
Negative skin friction occurs due to soil consolidating around the pile more than the axial
deformation of the pile itself- e.g.
-when fill is placed on top of soil adjacent to the pile subsequent to pile installation
-when the pile is installed after newly placed compressible fill, and the fill continues to
consolidate after pile installation.
Negative skin friction can cause extreme problems if a deep compressible layer (e.g. soft
peat) is overlain by gravel and a new fill is placed on top of the gravel: the peat compresses
and huge shear stresses are generated by the frictional gravel which drags the pile into the
ground and cause failure.
Negative skin friction can occur on the perimeter of a pile driven through a layer of clay
undergoing consolidation (e.g. due to a fill recently placed over the clay) into a firm bearing
stratum (refer figure below). The consolidating layer exerts a downward drag on the pile, and
therefore the direction of skin friction in this layer is reversed. The force due to this
downward or negative skin friction is thus an additional action (and factored accordingly)
instead of helping to support the external load applied at the head of the pile. It is
conservative to assume that the whole consolidating layer applies negative skin friction and
only the soil below contributes to the resistance of the pile, though in reality the cross- over
point between negative and positive skin friction (also called the neutral plane) will occur
within the consolidating soil, where the settlements of the soil and the pile are equal.
12
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
Where negative friction is anticipated, it must be included in the pile design: we must design
for a load which includes downdrag on the pile.
𝑄𝑢𝑙𝑡
𝑊𝐿 + 𝑁𝑆𝐹 = … … (2.16)
𝐹
Where F=2.5 without a pile test.
For underreamed piles without a pile test, the working load is calculated as the lesser values
given by the equation (2.16) above and (2.17) below:
𝑄𝑏
𝑊𝐿 + 𝑁𝑆𝐹 = 𝑄𝑠 + … . . (2.17)
3
13
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
The ultimate load capacity of a group piles will not necessarily be the same as the sum of
individual pile capacity which makes up the group. Similarly, the settlement of a pile group
will usually be very different from the settlement of a single pile carrying the same average
load.
The efficiency ɳ:
(Qg𝑟𝑜𝑢𝑝 )/n
ɳ= … . . (3.2)
𝑄𝑠𝑖𝑛𝑔𝑙𝑒
(i) Block failure: the soil contained within the pile group moves down with the piles
as block. The failure load of the block is given by:
𝑄𝑏𝑙𝑜𝑐𝑘 = 𝐴𝑠 𝑐𝑢 + 𝐴𝑏 𝑁𝑐 𝑆𝑐 𝑐𝑢 … … (3.3)
where
14
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
(ii) Individual pile failure: Each pile punches into the surrounding ground. However,
model tests have shown that even individual pile failure is occuring, the efficiency
ɳ is less than 1 due to the interaction between the piles.
1 1 1
2
= 2 + 2 … … . . (3.4)
𝑄𝑔𝑟𝑜𝑢𝑝 𝑄𝑏𝑙𝑜𝑐𝑘 (ɳ𝑄𝑠𝑖𝑛𝑔𝑙𝑒 )
The efficiency, ɳ given by (3.2) increases from a value of about 0.7 when the spacing factor
is about 2, to a value of 1 when the spacing factor is greater than about 8. This ia a reasonable
rule of thumb to use for design.
15
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
16
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
The pile group can be treated as an equivalent raft acting at the base of the piles. For a deep
clay layer underlying the piles, the area of the equivalent raft is given by a 1:4 spread. If the
stiffness of the soil increases significantly at or shortly below the base of piles, no spread of
load is assumed.
The equivalent raft at the bottom of the piles is taken to have the same dimensions as the pile
group. The Burland and Burbidge (1985) method for shallow foundations should then be
applied.
17
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
18
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
When piles are subjected to soil movement, these piles are known as passive piles. Soil
movement is encountered in practice when piles are placed in an unstable slope, landslides,
adjacent to deep excavation, tunnel operation, marginally stable riverbank with high
fluctuating water level and also in piles supporting bridge abutment adjacent to approach
embankments. The design of such piles may be based on the assumptions that forces from
moving soil will act against the piles and ‘squeeze’ past the piles. On the other hand, active
piles referred to a pile subjected to external horizontal force. Refer figure below for the
differences between active and passive piles.
Active Passive
19
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
1 + sin 𝜑′
𝐾𝑝 = … … . (4.2)
1 − sin 𝜑′
It is assumed that the limiting lateral soil pressure pu on pile rises linearly from 2cu at the
surface to 9cu at a depth of 3 pile diameter (z=3d).
20
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
7
𝑝𝑢 = (2 + ( ) 𝑧) 𝑐𝑢 𝑧 < 3𝑑
3𝑑
9𝑐𝑢 𝑧 > 3𝑑 … . (4.3)
Failure mechanism for a short free-headed pile, showing limiting earth pressures for
undrained clay.
21
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
By horizontal equilibrium:
22
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
H=Pab …..(4.7)
For a pile restrained at the head, a plastic hinge must form at the head as well. This is shwin
below. Mp should be added to the right hand side of the equation (4.8).
23
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
24
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
25
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
26
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
Qa 𝑄 sin 𝜃 ≤ 𝑄𝑎 → 𝑄 ≤ 𝑄𝑎 𝑐𝑜𝑠𝑒𝑐 𝜃
Q 𝑄 cos 𝜃 ≤ 𝑄1 → 𝑄 ≤ 𝑄1 𝑠𝑒𝑐 𝜃
Where
Ql
𝜃 = angle between Q and Ql;
The maximum possible load capacity occurs with the simultaneous mobilization of Qa and Ql:
27
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
28
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
29
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
30
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
31
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
32
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
33
University of Nottingham (Malaysia Campus)
Department of Civil Engineering H23GGE Foundation and Earthworks
34