Professional Documents
Culture Documents
Bengt H. Fellenius, DR - Tech., P.Eng
Bengt H. Fellenius, DR - Tech., P.Eng
Lecture 2
Wide pile groups
N.B., the above approach goes far beyond the 1948 Terzaghi and Peck
suggestion of placing an Equivalent Raft at the lower third depth
2
Epile+soil and Footprint Ratio, FR
The combined E-modulus of the pile and soil body, the equivalent pier, is
expressed in the equation in the previous slide. the E-modulus of the pile
material is either 200 GPa (steel piles) or about 30 GPa (concrete piles) and
the soil modulus is usually no more than about 50 MPa or smaller, often
much smaller. Therefore, the combined modulus depends mainly on the pile
modulus and the Footprint Ratio, defined as follows.
Epile +soil = FRxEpile + (1 – FR)Esoil ≈ FRxEpile [Esoil is very small compared to Epile]
QH
L
EPile S o il ARa ft
3
The response of the perimeter piles is a key issue to discuss.
The equivalent raft is a flexible raft, and so is,
approximately, also the foundation raft. The
perimeter piles will therefore carry a smaller
portion of the load than the interior piles. Should
the perimeter piles then not be shorter than the
interior piles to minimize differential settlement?
FILLS, etc.
Foundation raft
G.W.
FILLS, etc.
Foundation raft
G.W.
5
Indeed, for analysis and design of a piled foundation one must
address settlement. Addressing “capacity” of single piles in the
group and treating the “capacity” of the entire group as a
collection of single piles is not constructive, nor is it safe.
6
A wide pile group supporting a liquid storage tank in Tessaloniki, Greece (Savvaidis 2009)
20
Compression of 15
piles is minimal.
10
112 1.0 m diameter, bored The settlement
5
piles installed to 42 m depth
Norht-South
originated from
below the pile 0 Pile 16
diameter. -20
Dense, silty sand to 50+ m depth -20 -15 -10 -5 0 5 10 15 20
East-West
7
The settlement of the Tessaloniki tank was monitored during the hydro test and I have
used the settlement measured at the tank mid-point in a back-calculation, which
calibrated the soil parameters for analyzing the settlement across the tank diameter.
10
20
30
40
8
A second example: The Gdansk Millennium Bridge across the Dead Vistula River
22.4 m
#2 #3 #4
#1
22.4 m
#8 #5
#7 #6
52.4 m
SETTLEMENT (mm)
10
Land Side
15
Average
20 River Side
10
Settlement of the ground surface due to embankments placed on soft
compressible clay (data from Broms 1976). The pile-soil, i.e., the pier,
compresses less than the without piles (columns) and most of the measured
settlement was due to compression of the soil below the columns.
≈10 m
-10 -5 0 5 10 15
0
1
SETTLEMENT (mm)
10 2
1: 16 days
6m 2: 65 days
20 3
4 3: 351 days
4: 541 days
30
40 ≈9 m 3
50 Column Reference
Area (half) 4 Area (half)
60
11
Single Pile
Conventional "Group Effect" Concept applies
a “group efficiency coefficient”
“Group efficiency”
9-pile Group 800 usually refers to
Average of 4 Piles
“capacity” if a single
4-pile Group
PILE HEAD MOVEMENT (mm)
1,200
800
400
500 mm Pile Diameter
Distance: 3.0 m (#3)
0
0 4 8 12 16 20 24
MOVEMENT (mm)
Building 1, Building 2,
211 piles 104 piles
Soft to very soft, compressible, normally consolidated, marine clay extending to about 40 m depth. The general area was subject to
a small general subsidence, about a millimetre or two per year. The nominal total load on the foundations of the two buildings were
46 and 54 MN, respectively, corresponding to an average stress over the building footprint of 66 kPa and 60 kPa, respectively, i.e.,
very similar values.
Both buildings were supported on 18 m long about 300 mm diameter piles. However, the average axial working load was 220 kN/pile
for one building and 520 kN/pile for the other. The settlement and contact stress for the buildings were monitored over long time.
Building 1, Building 2,
211 piles 104 piles
Soft to very soft, compressible, normally consolidated, marine clay extending to about 40 m depth. The general area was subject to
a small general subsidence, about a millimetre or two per year. The nominal total load on the foundations of the two buildings were
46 and 54 MN, respectively, corresponding to an average stress over the building footprint of 66 kPa and 60 kPa, respectively, i.e.,
very similar values.
Both buildings were supported on 18 m long about 300 mm diameter piles. However, the average axial working load was 220 kN/pile
for one building and 520 kN/pile for the other. The settlement for the buildings was monitored over 13 years.
Average stress from the building weight over the building footprint was
almost same for the two buildings, i.e., 66 and 60 kPa, respectively).
Building 1 with twice as many piles as Building 2 settled less within the pile
length (less compression) and more below the pile toe level (due to the 10 %
larger stress). That is, the differences compensated each other.
PIER
W
Foundation
Piles
b = 406 mm Bored Pile Wall
D = 48 m b = 800 mm
D = 12.0 m
(To prevent scour.
Free from contact
with raft and piles).
19.0 m
10.6 m
The soil profile consisted of about 10 m of clay on about 10 m of dense sand underlain by soft clay deposited at about 48 m
depth on a very dense sand and gravel bed. The clay is normally consolidated undergoing small. The piled raft foundation
comprised 144 mandrel-driven, then concrete-filled, steel pipe piles, 406-mm diameter, 48 m long, uniformly distributed in a
10.6 m by 19.0 m raft and driven into the very dense sand and gravel layer. The pile c/c distance was 1.2 m (3.0 pile
diameters). The Footprint Ratio was 9 %. The unfactored load from the pier was 800 kN/pile, which incorporated a factor of
safety of 3.0 on pile capacity as stated to have been verified in static loading tests.
PIER
W
800
CORNER
0
19.0 m
10.6 m As should be for a raft on elastic soil, the load on the perimeter piles was larger
than that on the interior piles with the load on the corner piles larger than that
on the side piles. With time, the subsiding soil created downdrag and the
perimeter piles settled and transferred load to the interior piles.
The soil profile consisted of about 10 m of clay on about 10 m of dense sand underlain by soft clay deposited at about 48 m
depth on a very dense sand and gravel bed. The clay is normally consolidated undergoing small. The piled raft foundation
comprised 144 mandrel-driven, then concrete-filled, steel pipe piles, 406-mm diameter, 48 m long, uniformly distributed in a
10.6 m by 19.0 m raft and driven into the very dense sand and gravel layer. The pile c/c distance was 1.2 m (3.0 pile
diameters). The Footprint Ratio was 9 %. The unfactored load from the pier was 800 kN/pile, which incorporated a factor of
safety of 3.0 on pile capacity as stated to have been verified in static loading tests.
9.4 m
10.6 m
1.5 m
Monitored
Piles
1.5 m 0.7 m
Single Pile
LOAD (kN)
-1,000 0 1,000 2,000 3,000 4,000 5,000
9.4 m
0
The working load?
10
43 m long
single pile
installed
away from
10.6 m
group
20
DEPTH (m)
1.5 m
30
Monitored
Piles 40
Interior
piles
1.5 m 0.7 m
50
The result showed that interior piles did not develop any drag force—nor any shaft resistance. The toe force
was about equal to the applied load per pile. The perimeter pile responded the same way as the single pile.
LOAD (kN)
-1,000 0 1,000 2,000 3,000 4,000 5,000
9.4 m
0
The working load?
10
43 m long
single pile
installed
away from
10.6 m
group
20
DEPTH (m)
1.5 m
( )
30
Perimeter
pile
Monitored
Piles 40 Perimeter pile
Interior with loads
piles shifted to same
1.5 m 0.7 m load at start as
the others
50
Single Pile
Contact
Contactstress
stress
Moreover, most case histories on pile group observations report only short-term
observations and often only the load-movement of the pile head (or of the supported
raft). Only few cases exist that report also the contact stress. I know of none reporting soil
stress in-between the piles versus depth.
23
Comments on the conventional model
Piled Foundation Soil If there is contact between raft and soil, there is contact stress and
there is axial pile force, but there is no movement between pile and
? ? soil. Therefore, the first fact is that under the pile cap (raft), the strain
in the pile and in the soil must be equal and no relative movement can
? ? ? develop between the pile and the soil.
The second fact is that two interacting bodies must have strain
compatibility. But for the very surface of the combined body of pile
and soil, the strain is equal across the body. Therefore, there cannot be
any relative movement between the two parts and there cannot be any
shaft resistance.
The third fact is the E-modulus of the pile and of the soil determine the
magnitude of axial stress in the pile and vertical stress in the soil.
The fourth fact is that right below the raft, the soil is usually an
engineered soil with large E-modulus and, therefore, the contact stress
is large. Deeper down , the soil is usually soft with small E-modulus.
The total load is the same and the soil stress is therefore smaller and
For the interior the axial stress in the pile larger. The strain across the body of piles and
piles, this model is soil will have increased because the combined E-modulus has
simply not true! decreased. But this occurs without any shaft resistance developing.
At some shallow depth below the raft, distribution of stress and strain
Think of the pile and soil as a loaded reinforced concrete
in piles and soil is independent of the magnitude of contact stress and,
column: The distribution of stress in the concrete and
stress in the rebars is governed by the areas of concrete indeed, whether there at all is a contact stress between the raft
and steel and the particular E-moduli of the concrete and underside and the soil surface.
steel. Moreover, if the concrete would be chipped off
along some length, all load will be in the steel rebars If there is ongoing consolidation of the soil inside the pile group, some
along that length, but above and below, strain amount of drag force may develop above a neutral plane somewhere
compatibility still controls what stress there is in the along the pile length. The drag force cannot be larger than the buoyant
concrete and what there is in the steel. The “chipped- weight of the soil in-between the piles above that neutral plane.
off” length would be like a layer of very soft soil.
24
Correct condition of shear between an
interior pile and soil
Supported load Interior piles in a group engage the soil from the
(shown as stress)
pile toe upward in contrast to single piles which
engage the soil from the pile head downward
Contact Stress (the Franke principle).
per strain
compaibility for The pile toe movement (downward) is of course
pile and soil
Apparent equal to the “upward” movement of soil in-between
upward soil the piles. The movement is resisted by shaft
movement Shaft resistance resistance along some length above the pile toe.
Pile toe only near the pile
penetration toe level.
The load applied to the pile from the raft is
Toe Stress transferred down to the pile toe, forcing the pile toe
downward and, thus, generating shaft resistance
above the pile toe that reduces the axial force in the
pile until an equilibrium is reached between the pile
The shaft resistance is determined by toe force and the axial force reduced due to the
t-z relation (for the pile element at the
pile toe) and the toe stress (force) is
movement-generated shaft resistance.
determined by the q-z relation.
25
Compare “Plugging” of an Open-toe Pipe Pile
Forces on pipe in dynamic Forces and Movement
and static loading of the core in static loading
(The pipe is not shown)
Movement
Core between core
Core
and inside
of pipe
Core
Core
26
The Franke principle is supported by
measured load distribution for interior piles
Kakurai, M., Yamashita, K., and Tomono, M., 1987. Settlement
Yamashita, K., Wakai, S., and Hamada, J. 2013. Large-scale piled
behavior of piled raft foundations on soft ground. Proceedings of
raft with grid form deep mixing walls on soft ground. Proc. 18th
the 8th Asian Regional Conf. on SMFE ARCSMFE, Kyoto, 20 -24 July
ICSMGE, September 2 6, Paris, France, Vol. 3, pp. 2637-2640.
1987, Vol. 1. pp. 373 -376.
LOAD (kN)
LOAD (MN) & SPT-N (blows/0.3m)
0 5 10 15 20 0 100 200 300 400 500 600 700 800
0
0 400 mm diameter, Raft Level
closed-toe pipe-pile to
5 Raf t load 24 m depth, c/c = 4.5 m
Raft level per pile
10 5
End of One year
Construction later
15
Silt + Sand
20 10
DEPTH (m)
DEPTH (m)
Normally
consolidated
25 clay and silt
30 15
Reduction of axial
35 load due to shaft
resistance above Alt. 1
the pile toe
40 20
45 Alt. 2
PILE
50
25
F
E
2 2
E
D D
4 4
C
C
B
6
B
6
A
The simulated
load distributions
A 1 2 3 4 5 6
8 8
agree well with
DEPTH (m)
DEPTH (m)
FF
22 22
EE
DD
44 44
CC
BB
66 66
AA
11 22 33 44 55 66
88 88
DEPTH (m)
DEPTH (m)
Interior
Interior
10
10 10
10 Piles
Piles
12
12 12
12
Perimeter
PerimeterPiles
Piles
14
14 14
14
Pile
PileToe
Toe Depth
Depth
Pile
Pile 16
16
16
16
Soil
Soilat
at Movement
Movement
Perimeter
PerimeterPile
Pile
18
18 18
18
Soil
Soilat
at
Interior
InteriorPile
Pile
20
20 20
20
100 4.0
0 0.0
0 2 4 6 8 10 12 14 16 18 20
The blue curve is the pile-toe force vs. the pile-toe penetration. The green curve is the
sustained load on the pile (here 700 kN) minus the toe resistance for the series of pile
toe penetration. The toe penetration determines the length with shaft resistance of the
pile above the toe.
The right graph provides a check, showing the length of pile engaged by the balance
between the 700-kN sustained load and the 350-kN toe resistance mobilized by the
4.3-mm toe penetration (that both loads are 350 kN is just a coincidence)
400 20
33
Comparison of the pile–head load-movement response of
perimeter and interior piles for rigid and flexible rafts
Useful when estimating the load difference between
perimeter and interior piles
5,000
Rigid Raft, so
LOAD AT PILE HEAD (kN)
Slightly exaggerated,
qualitative graph 34
According to Horikoshi and Randolph 1997, the raft to soil stiffness ratio, Kr , across a raft loaded
by a uniform stress is governed by the below equation.
The stiffness ratio governs the Bending moment of a raft (mat), Mr, and differential
settlement, Δs, [M ≈ function of (Kr) and Δs ≈ function of (1/Kr)]. (Applies to rectangular raft).
For piled rafts, the Poisson Ratios are almost the same for pile and soil (Basile 2019).
The ratio between Er and Es is essentially equal to the Footprint Ratio, FR. Thus, if B≈L,
the relation for Kr becomes:
6 t 3
Kr
≈ ( )( )
FR L
35
For a center-to-center spacing, c/c, ranging from 5b through 3 b, FR ranges from
about 4 % through about 10% and Kr ranges from 0.09L through 0.12L, i.e. ≈ 0.1L.
0.1
0.10
0.1
0.10
Thus, most piled rafts can be considered at least “practically flexible”.
Yet, for a uniformly loaded raft, perimeter piles will still receive larger loads than
the interior piles. In case of general subsidence at the site, in the long-term, the
difference reduces; perimeter piles may even receive smaller loads.
36
Case History: Full-scale tests at the Bolivian B.E.S.T. site in Santa Cruz on a
group of thirteen 300-mm piles (FDP) constructed to 9.5 m depth as FDP in
silty sand and equipped with a bidirectional cell at the pile toe.
Sequence of EB Expansion
1st: E8, E2, 3.45 mand E14
E4, E12,
Phase 1b. Bidirectional load vs. upward movement (no pile cap)
2nd: E3, E9,
2.54 mand E7
E13,
3rd: E11, E10, E5, and E6
E2 E3 E4
E5 E6
3.45 m
E8
E7 E9 Plan View All pile loads were the same (all BDs were connected to the same
hydraulic system). The upward pile movements were separately
E10 E11
measured for each pile.
E12 E13 E14
FRpile = 14 %
` ` Vertical view
Bidirectional
cells (BD)
600
E8
BD LOAD (kN)
E9 500 E-Group
E7
400 Center
E10 E11 Interior
Mid-side
300
E12 E13 E14 Corner
200
FRpile = 14 %
100
0
0 5 10 15 20 25 30
MOVEMENT (mm)
Sequence of EB Expansion
1st: E8, E2, E4, E12, and E14
25
UPWARD MOVEMENT (mm)
E5 E6
20 E8
E9
435 kN
E7
E10 E11
420 kN
15 E12 E13 E14
404 kN
10
Bidirectional 391 kN
5
cells (BD) 346 kN
0
0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6
DISTANCE ALONG DIAMETER (m)
E8
E7 E9
E10 E11
E12 E13 E14
39
Load measurements for interior piles in the E-group test
Load as
SOIL Soil Stress The distribution of axial force for the
interior piles showed agreement with the
Unified method applying the Franke
principle: the pile shaft resistance was
BIDIRECTIONAL engaged from the toe upward.
CELL
Also in the head-down test, the pile
EXPANDER
BASE Load at Pile Shaft Resistance group responded as a single pier.
Toe (in BD) above Pile Toe
40
Load-Movement Results
Contact Load
1,000 1,000
E-GROUP Total Load
Average
Load All Piles
600 600 Average
Interior Piles
SHAFT Sequence of EB Expansion
1st: E8, E2, E4, E12, and E14
Resistance 2nd: E3, E9, E13, and E7
400 400 3rd: E11, E10, E5, and E6
E2 E3 E4
Average E5 E6
Contact Load E8
200 200 E7 E9
E10 E11
E12 E13 E14
0 0
0 10 20 30 40 50 0 10 20 30 40 50
HEAD MOVEMENT (mm) PILE-HEAD MOVEMENT (mm)
41
A current design project in Santa Cruz de la Sierra, Bolivia
Small group
Wide group
Narrow group
5 5 5 5
DEPTH (m)
10 10 10 10
15 15 15 15
DEPT H (m)
EB
Ran ge of pile
2020 0 25
20
50 bl/0.3m
20 20
toe levels
determined
fr om the
assigned
w or king loads EB
40
40
44
60
WEST TOWER EAST TOWER
50
NORTH-SOUTH (m)
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130
EAST-WEST (m)
45
Settlement calculations in UniSettle --- Input
WEST TOWER
46
Settlement calculations in UniSettle --- Output
Numbers are distorted because the two
rafts, Rafts A and B, are typical, only.
================================== ==================================
Settlement
distribution
47
CONCLUSIONS
1. The analysis of how a pile group responds to an applied load start with analyzing the response of a
single pile because perimeter piles will behave much like a single pile—both in regard to supporting
an applied load and downdrag effect. One difference is that the load placed on the single pile is
known, but what load that will be on the perimeter piles depends on a number of factors, such as
number of piles, rigidity of the raft, the distribution of settlement below the pile toe level, and
amount of downdrag.
2. There will be contact stress under a raft resting on soil. However, the distribution between load
acting on the piles and what load is represented by the contact stress across the ‘free’ area
depends on the E-moduli of pile and soil, and on the footprint ratio. Strain compatibility rules for
pile stress and soil stress. Some distance below the raft, there is no difference whether or not there
is contact stress under the raft. Thus, there is no “contribution to bearing” from contact stress.
3. The axial force in the piles due to the applied load will not reduce until near the pile toe level—a
boundary between pile body (pier) and soil—As the pile toe is pushed into the soil, toe resistance
builds up at the same time as shaft resistance is mobilized above the pile toe. An equilibrium
(balance) is attained when the movement has resulted in shaft and toe resistances that together
are equal to the axial pile force from above.
4. Knowing the load-movement response of a pile is very important for the analysis, However, a
“capacity”, however defined, of the individual piles in the group, or of the group as a whole, is
irrelevant to the design. The controlling aspect is the settlement of the piled foundation (average
and across the raft), which is governed by the compression of the soil below the toe level,
compression of the piles (pier effect), load distribution across the raft (raft rigidity), and load-
transfer analysis for the pile toe.
Thank you for your attention
49