1

Lecture 3
Settlement of Single Pile and
Pile Group
Associate Professor C F Leung
Department of Civil Engineering, NUS
Email : [email protected]
Kaohsiung Taiwan Aug 2004
Building settles excessively but rather uniformly due to
water drawdown caused by nearby subway construction.
1
st
storey now becomes the basement!!
2
Differential settlement is often a major concern.
Pisa Tower is the famous case.
Tilt of building under construction in
Singapore, Jan 2003
Tilt angle 0.1
o
Top of building displaces 0.3048 m
Thus settlement can be the
governing factor in pile design!
3
3.1 Settlement of Axially Loaded
Single Piles
Elastic theory is used to compute settlement.
Pile-settlement behaviour can be obtained
through dimensionless parametric solution
based on pile and soil properties (Poulos and
Davis).
Concept is shown in Fig. 3.1
Fig. 3.1
Analysis of
floating
pile (pile
base not
resting on
rock)
[After
Poulos and
Davis]
4
Elastic Theory
Pile is divided into a number of elements (Fig. 3.1(e))
The vertical displacement (s) of the soil adjacent to
the pile at element i resulting from a stress P
j
on an
element j is given by
s
ij
= (d/E
s
) I
ij
P
j
where d = pile diameter
E
s
= Youngs modulus of soil
I
ij
= vertical displacement factor for element i due
to shear stress at element j
=

I
=
After Mindlin (1936)
[Refer to Poulos and Davis Appendix A]
z
1
, z, R
1
, and R
2
are some physical distance between
element i and element j
The soil displacement at element i due to P of all the
elements and the base is
s
i
=
where d
b
= diameter of pile base
P
b
= stress at pile base
5
The displacement of the pile elements is taken care of
by the pile stiffness factor, K = E
p
R
A
/E
s
where E
p
= Young modulus of pile material
R
A
= pile area ratio = A
p
/(d
2
/4)
= 1 for pile with solid cross-section.
By means of finite difference method and by ensuring
that the stress and displacement are compatible for all
the elements, solutions for stress and displacement
along the pile are obtained.
[The most relevant one is the displacement (settlement)
at the pile top]
3.2 Floating Piles
Piles with base not resting on rock (hard stratum) i.e. Youngs
modulus of soil at pile base E
b
= E
s
.
Pile head settlement
= P I / (E
s
d)
where P = applied axial load on pile
I = influence factor
= I
0
R
k
R
h
R
v
(Figs. 3.2 to 3.5)
d = pile shaft diameter
and E
s
= Youngs modulus of soil
6
Fig. 3.2
Settlement
influence factor
I
0
for
incompressible
pile in soil with
Poisson ratio
s
= 0.5 [After
Poulos and
Davis]
L = pile length
d
b
= diameter
of pile base
Fig. 3.3 Compressibility correction factor for settlement, R
K
Pile stiffness
factor
K = E
p
R
A
/E
s
where
E
p
= Youngs
modulus of pile
material;
R
A
= area ratio
= A
p
/(d
2
/4)
= 1 for solid pile
[After Poulos and Davis]
7
Fig. 3.4 Depth correction factor for settlement R
h
h is the
thickness of
soil to hard
rock (rigid
stratum)
[After
Poulos and
Davis]
Fig. 3.5 Poissons ratio correction factor for
settlement, R

[After Poulos and Davis]

8
3.3 End Bearing Piles
Pile with base resting on hard rock end bearing
stratum, i.e. E
b
> E
s
.
Refer to Fig. 3.6 for the analysis of end bearing pile.
Pile head settlement
= P I / (E
s
d)
where I = I
0
R
k
R
b
R
v,
and R
b
= correction factor for stiffness of bearing
stratum (Fig. 3.7)
Fig. 3.6
Analysis of
end-bearing
pile
Mirror image is assumed at the
hard bearing stratum
[After Poulos and Davis]
9
Fig. 3.7 Base modulus correction factor for settlement, R
b
Fig. 3.7 (cont.)
E
b
=
Youngs
modulus of
soil at pile
base
[After Poulos and Davis]
10
3.4 Non-Homogeneous Soil
If modulus variation between successive layers along
pile length L is not large, use average soil modulus
E
av
= [(E
i
h
i
)]/L
where E
i
= modulus of soil layer i,
h
i
= thickness of soil layer i.
For a pile penetrates one soil layer and founded in
the second distinctly different soil layer:
(1) Treat the portion of pile in 1
st
layer as end bearing
pile, determine settlement, s
1
of this portion
subjected to applied load P
(2) Determine the load transfer at the interface of the 2
soil layers P
2
.
(3) Calculate settlement in 2
nd
layer, s
2
subjected to P
2
.
(4) Sum up s
1
and s
2
to obtain overall settlement of
pile head.
11
3.5 Effect of Underlying
Compressible Strata
Settlement due to compressible strata below pile base
must be considered in calculating pile settlement.
For pile founded within the first layer of a system
consists of m layers of different soils and the depth
of the first layer, h
1
> L, pile settlement
pile settlement in
founding layer
displacement of the
underlying layers
caused by the pile
where
o
= settlement of a floating pile in a layer of soil
of depth h
i
(see Section 3.2),
I
j
= displacement influence factor I

on the pile
axis at the level of the top of soil layer j
(Fig. 3.8),
and E
sj
= Youngs modulus of soil layer j.
12
Fig. 3.8
Influence
factors for
settlement
beneath centre
of a pile
[After Poulos and Davis]
3.6 Effect of Pile Cap Resting on
Soil Surface
Previous solutions derived based on free standing
piles with pile cap not in contact with soil.
Settlement is reduced if pile cap is resting on the soil.
However, the effect is only significant for relatively
short piles (e.g. L/d 10).
13
3.7 Soil Parameters
(1) Clay
Backfigured drained Youngs modulus E
s
for pile
in clay, refer to Fig. 3.9.
Drained Poissons ratio v
s
as follow:
Stiff OC clays 0.1 to 0.2 (ave 0.15)
Medium stiff clay 0.2 to 0.35 (ave 0.3)
Soft NC clays 0.35 0.45 (ave 0.4)
Undrained soil parameters (for immediate
settlement calculation only, therefore not so useful)
E
u
= 1.5 E
s
/(1+ v
s
), and
v
u
= 0.5 for saturated clay
Fig. 3.9 Backfigured
soil modulus E
s
for
piles in clay (after
Poulos and Davis)
Note differences in
E
s
values (including
upper limit of E
s
)
for driven and
bored piles
14
(2) Sand
Ultimate settlement occurs immediately upon loading
Use drained soil parameters, E
s
and v
s
to calculate
pile settlement.
Suggested average values of E
s
along shaft, refer to
Table 3.1.
Soil modulus at pile base, E
b
> average E
s
along shaft
Poulos suggested an upper limit of E
b
= 10 E
s
for
driven piles in dense sand, and a lower limit of E
b
= 5
E
s
for bored piles in loose sand.
With absence of test data, use v
s
= 0.3.
Table 3.1 Suggested average values of E
s
for driven piles
in sand (after Poulos and Davis)
15
3.8 Settlement of Axially Loaded
Pile Groups
Simplified equivalent raft approach
- Soil is treated as continuum and presence of piles is
ignored.
- Load is transferred to a depth where soil is stronger
and less compressible (Fig. 3.10)
- Total pile group settlement = settlement of virtual
foundation placed at some depth below ground
surface.
- Final settlement = (immediate settlement +
consolidation settlement) of the virtual foundation.
Fig 3.10 Settlement of Pile Groups (Equivalent raft concept)
16
Fig 3.10 cont. (After US Navy Design Manual DM7)
3.9 Interaction Factor Approach
Consider two identical equally loaded piles (Fig.
3.11) in an elastic mass to obtain relationship between
centre-to-centre spacing between piles, s and
interaction factor, .
= ratio of additional settlement caused by adjacent
pile to settlement of pile under its own load.
17
Fig 3.11
Group of two
floating piles
[After Poulos and Davis]
A load on Pile 1
would cause
additional settlement
on Pile 2 and verse
versa.
The total pile group
effect is determined
using Principle of
superposition.
3.10 Floating Piles
Interaction factor
F
, for two floating piles in a
homogeneous semi-infinite mass having v
s
= 0.5 is
dependent upon
(1) Dimensionless pile spacing s/d,
(2) Pile length/diameter ratio, L/d
(3) Pile stiffness factor, K ( K= E
p
R
A
/E
s
)
Refer to Fig. 3.12 for
F
values for L/d = 25.
Correction factor N
h
for finite layer depth (Fig. 3.13)
and N
db
for piles with underream (Fig. 3.14).
18
Fig 3.12 Interaction factors
F
for floating piles, L/d = 25
Refer to Poulos and Davis on charts for other L/d values
[After Poulos and Davis]
Fig 3.13
Correction
factors N
h
to
interaction
factors, for
effect of finite
layer depth
(L/d = 25)
[After Poulos and Davis]
19
Fig 3.14 Correction factors N
db
to
interaction factors, for effect of d
b
/d
[After Poulos and Davis]
Fig 3.15 shows the
correction factors
N

for effect of
Poissons ratio
s
for L/d = 50, K =
1000 and h/L = .
[After Poulos and Davis]
Variation in
s
has
relatively insignificant
effect on , except for
relatively large s/d
values. As an example,
20
3.11 End Bearing Piles
Selected values of interaction factors
E
for end
bearing piles on a rigid stratum are shown in Fig.
3.16.
For piles resting on not-so-rigid stratum, values lie
between
F
(for floating piles in homogeneous mass)
and
E
(for end bearing piles resting on rigid base)
and can be expressed as
=
F
F
E
(
F
-
E
)
where F
E
= factor depending on K, L/d, and E
b
/E
s
,
see Fig. 3.17.
Fig 3.16 Interaction factors
E
for end-bearing piles, L/d = 25
[After Poulos and Davis]
21
Fig 3.17 Interaction reduction factor F
E
for L/d = 25
[After Poulos and Davis]
3.12 Pile-Raft Foundation
Solutions presented earlier are derived based on
freestanding pile group (pile cap not resting on soil).
For pile-raft foundation with the pile cap resting on
soil, selected values of interaction factors
r
for a
pile-raft unit is given in Fig. 3.18.
In the example shown in Fig. 3.18, the piles are
incompressible (i.e. K = ) with L/d = 25, pile cap is
rigid and v
s
= 0.5. Other solution charts are given in
Poulos and Davis.
22
Fig 3.18 Interaction factors
r
for pile-raft units with L/d = 25
[After Poulos and Davis]
d
c
= diameter of pile cap
corresponding to 1 pile
For pile-raft systems with several pile cap units, each
having an equivalent value of d
c
/d such that the area
occupied by the unit is the same as that occupied by a
typical portion of the cap in the group.
For example, for a square arrangement of piles in the
group,
equivalent d
c
/d = (4/) s/d.
- For a pile-raft foundation, the settlement of single
pile is reduced as part of the load is transferred to the
soil but the pile cap-soil-pile interaction increases as
shown in Fig. 3.18.
23
3.13 Other considerations
(a) Soil with increasing modulus with depth

(soil with increasing modulus with depth)

75 to 80%
(homogeneous soil)
(b) Group with different pile spacing and size
For two piles of different size, the increase in settlement of
pile i caused by pile j (Fig. 3.19),

ij
= p
j

ij
where
j
= settlement of pile j under its own load
and
ij
= interaction factor corresponding to the
spacing between piles i and j, and for the
geometrical parameters (i.e. L and d) of pile j.
[After Poulos and Davis]
Fig. 3.19
24
(c) Effect of Underlying Compressible Strata
Extend method in Section 3.5 for pile group
calculation.
Overall settlement = settlement of the group in
founding layer + settlement of the underlying layers
caused by all piles in the group.
Pile group can be assumed to be an equivalent single
pier. (e.g. of the same gross plan area and of
equivalent length L
e
)
Refer to Poulos and Davis for details.
3.14 Analysis of General Groups
Use superposition of two pile interaction factors to
analyse settlement behaviour of a pile group.
Thus, a group of n identical piles, settlement of any
pile k in the group,
k
is given as
where
1
= settlement of single pile under unit load
P
j
= load in pile j, and

kj
= interaction factor for spacing between
piles k and j.
25
For group containing different sizes or geometry of
piles,
where
ij
= settlement of single pile j under unit load
and
kj
= interaction factor for spacing between
piles k and j, and for geometrical
parameters of pile j.
For vertical load equilibrium, the total pile group, P
G
is
given by
For a pile group of n piles, there will be n displacement and
one load equations, these can be solved for two simple
conditions:
(1) Flexible pile cap
- equal load or known loads on all individual piles.
- all P
j
are given to solve settlement of individual piles
j
Hence the differential settlement among piles can be
evaluated.
26
(2) Rigid pile cap
Equal settlement for all piles i.e. all
j
are equal
Only load on the whole pile group P
G
is given
Distribution of load on individual piles in the pile
group is not uniform (unless symmetrical).
For practical purpose,
average settlement of group with equally loaded piles
settlement of pile group with rigid cap
Suggestion: Need not remember how to use the
equation of summation, always work from first
principle.
Example 3.1
Use the same data as Example 2.1.
(a) Two number of 0.8 m diameter bored piles with
penetration depth of 31 m are adopted.
From Fig. 3.9
Depth (m) Soil Av. c
u
(kPa) E
s
(MPa)
0 - 5 Firm silty clay 50 6.5
5 25 Stiff silty clay 100 22
25 - 31 Hard silty clay 500 80
27
Average E
s
along pile shaft of 31 m (should actually
use the more rigorous approach as the soil modulus
along pile varies greatly)
E
s
= (5 x 6.5 + 20 x 22 + 6 x 80)/31 = 30.7 MPa
E
b
= 80 MPa
E
b
/E
s
= 80/30.7 = 2.6
L/d = 31/0.8 = 38.75
K = E
p
R
A
/E
s
= (28 x 10
3
) x 1/(30.7) = 912
E
p
(= E
conc
) value assumed.
Pile settlement per unit applied load for a single
pile = I/(E
s
d)
Based on parameters established earlier:
Fig. 3.2 gives I
o
= 0.053
Fig. 3.3 gives R
k
= 1.22
Fig. 3.5 gives R

= 0.93 (assume
s
= 0.3)
Fig. 3.7(b) and (c) give R
b
= 0.92 by interpolation
Thus = (0.053x1.22x0.93x0.92)/(30.7x0.8)
= 0.00225 mm/kN
28
Pile group
Try s/d = 2 to start with
Use L/d = 25 charts (should actually interpolate from
L/d = 25 & L/d = 50 charts for L/d = 38.75)
Based on parameters established earlier,
Fig. 3.12 gives
F
= 0.58
Fig. 3.17 gives
E
= 0.06 {much smaller than
F
)
Fig. 3.18 gives F
E
= 0.2
Thus = 0.58 0.2(0.58 0.06) = 0.476
{interaction between piles is significant!}
(i) Flexible pile group
Working load on each pile = 6000/2 = 3000 kN
Owing to symmetry, settlement of each pile
i
is the same =
(0.00225 x 3000) mm x (1 + 0.476)
= 6.75 mm (1 + 0.476) = 9.963 mm
(1 is settlement due to loading on own pile and 0.476 is the
increase in settlement due to loading on the other pile)
Normally the settlement under working load should be less
than 10 mm. Thus the estimated settlement is within limit!
(ii) Rigid pile cap
Owing to symmetry, the load on each pile is the same at 3000 kN
and hence the settlement is also the same as the flexible pile
cap case.
29
(b) Precast RC piles
6 number of 0.32 by 0.32 m precast RC piles were used
(refer to Examples 2.1 and 2.4)
The settlement estimation procedure is essentially
similar to that for (a) bored piles except that
E
s
for bored piles and driven piles are different
for the same soil (due to installation effect)
Pile group
Unlike the previous 2 bored-pile group case which is
rather simple as there is only 1 interaction between 2
piles. For the 6 RC-pile group, each pile will interact
with 5 other piles with different s/d. Assuming
interaction between piles 1 and 2 is the same as that
between piles 2 and 1, there will be a total of 15 (5 +
4 + 3 + 2 + 1) interaction among piles. Thus the
calculation is tedious. Geometrical symmetry can
greatly reduce this tedious process.
30
Flexible pile cap
Taking load on each pile is the same at 1000 kN,
settlement of piles is not uniform.
Owing to symmetry, the 4 outer piles would have the
same settlement and the same settlement for the 2
inner piles.
Thus one only need to determine the settlement of 1
outer pile and 1 inner pile. The inner pile would
suffer higher settlement due to larger pile interaction.
Rigid pile cap
Settlement of all piles is uniformbut the load on individual
piles is not the same. Total load on whole pile group is the
same.
Owing to symmetry, the 4 outer piles would carry the same
load and the 2 inner piles would carry the same load.
Hence there are 3 unknowns: (a) pile settlement; (b) load on
outer pile and (c) load on inner pile.
Need to establish 3 simultaneous equations (1 on vertical force
equilibrium and the other 2 on the settlement of outer and
inner piles based on pile group interaction factors,
respectively).
To maintain uniform pile group settlement, the outer piles will
carry more loads.
31
3.15 Discussions on Elastic Theory
Limitations of Elastic Theory
Youngs modulus of soil (working load)
Pile group interaction (theory versus practice)
Pile centre-to-centre spacing
Floating piles versus end bearing piles
Rigidity of pile cap (thickness and plan size of cap)
Pile raft foundation
Recommendations for pile design in practice
Granite
Granite (Hillview area, Singapore)
Building
32
Buildings on widely varied ground
The elevation of the base of a proposed building is
shown in the previous slide. One end of the building
will be on ground with relatively shallow bedrock
(thus short piles) while the end will be on ground
with much deeper bedrock (long piles necessary).
Discuss the problems associated with the design and
construction of the foundation for the building.
Provide suggestions how the problems can be
overcome.