Sie sind auf Seite 1von 18

International Journal of Geotechnical Engineering

ISSN: 1938-6362 (Print) 1939-7879 (Online) Journal homepage: http://www.tandfonline.com/loi/yjge20

Numerical investigation of the lateral response of


battered pile foundations

Lassaad Hazzar, Mahmoud N. Hussien & Mourad Karray

To cite this article: Lassaad Hazzar, Mahmoud N. Hussien & Mourad Karray (2016): Numerical
investigation of the lateral response of battered pile foundations, International Journal of
Geotechnical Engineering, DOI: 10.1080/19386362.2016.1224030

To link to this article: http://dx.doi.org/10.1080/19386362.2016.1224030

Published online: 25 Aug 2016.

Submit your article to this journal

Article views: 26

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=yjge20

Download by: [Nipissing University] Date: 14 September 2016, At: 06:31


Numerical investigation of the lateral response
of battered pile foundations
Lassaad Hazzar1*, Mahmoud N. Hussien1,2 and Mourad Karray1
Three-dimensional finite difference parametric analyses have been carried out to investigate the influence
of vertical loads on the lateral performance of battered piles in both sandy and clayey soils. Different batter
angles and soil stiffness were considered to examine the salient features of this complex soilstructure
interaction problem. Numerical results show that the lateral response of battered piles embedded in
sands is influenced by the value and the direction of the pile inclination, the presence of vertical loads as
well as the soil state of density. When the lateral load acts in the opposite direction of the pile inclination
(negative batter angle), the lateral response of the piles increases substantially with the batter angle as well
as the sand packing. The response of piles at positive batter angles, however, does not appear to vary
considerably with the batter angle or the sand density. The effect of the vertical loads on lateral response
of battered piles in sands is found to be very pronounced. On the other hand, the lateral response of
piles embedded in a clayey soil at negative batter angles increases greatly with the inclination angle and
does not vary with its undrained shear strength. The lateral response of the piles with positive inclination
angles is independent of the batter angle and the soil stiffness. Moreover, the presence of a vertical load
prior to the application of a lateral load to a battered pile does not alter its lateral capacity.
Keywords: Battered piles, Lateral response, Vertical loads, Mohr circle, Finite difference modelling

Introduction inclined piles can be classified depending on their direction of


inclination with lateral loads to: Pile battered reverse, if the
Piles are used as a common foundation solution for high-rise lateral load acts opposite to the direction of pile inclination
buildings, high retaining walls, offshore structures, etc., and (negative batter angle), and Pile battered forward, if the lat-
are normally subjected to lateral loads in addition to their own eral load acts in the direction of the pile inclination (positive
vertical loads. When the horizontal load per pile exceeds what batter angle). General experimental observations reveal that
can be withstand by a vertical pile, battered piles are used in pile battered reverse offer more resistance than pile battered
conjunction with vertical piles to improve the overall efficiency forward. Lu (1981) concluded that the lateral capacity of pile
of the pilesoil system as they transmit the applied lateral loads is zero for a pile battered forward and maximum for a pile
partly in axial compression, rather than through shear and bend- battered reverse, indicating that the upper layer soil support
ing when only vertical piles are used. in a negative batter is enormous. In the same context, Zhang
Although the obvious superiority of battered piles in resist- etal. (1999), based on centrifuge test results and data reported
ing lateral loading, they do not receive an adequate geotechnical in the literature, reported that the lateral pile response increases
research interest compared to vertical piles. The most prom- over plumb piles were 4, 14, 24 and up to 50% in very loose,
inent research on this topic were some laboratory tests per- loose, medium dense and dense sands, respectively, at negative
formed on model inclined piles to investigate their behaviour angle batter (14). In contrast, the lateral response decreases
(e.g. Murthy 1965; Alizadeh and Davisson 1970; Meyerhof and over plumb piles were 4, 5, 15 and up to 35%, respectively, at
Ranjan 1973; Awad and Ayoub 1976; Ranjan et al. 1980; Hanna positive angle batter (14). Zhang et al. (1999) concluded that
and Afram 1986; Veeresh 1996; Zhang et al. 1999; Zhang et al. the effects of pile batter were significant in medium dense and
2002). These studies inferred that the values of batter angles dense sands, but minor in very loose and loose sands. The axial
as well as the direction of piles inclination with respect to the ultimate shaft resistance, however, decreases with an increase
applied lateral loads are the underlying factors that control the in the pile inclination as reported by Hanna and Afram (1986)
battered piles performance. According to Zhang et al. (2002), based on an experimental investigation using mode1 piles
pushed in medium dense sand deposits at different inclination
1
Department of Civil Engineering, Faculty of Engineering, Sherbrooke
University, Sherbrooke, Canada up to 30 with respect to the vertical and tested under vertical
2
Department of Civil Engineering, Faculty of Engineering, Assiut University, loads. Hanna and Afram (1986) attributed this reduction to the
Assiut, Egypt reduction in the average mobilised angle of friction between
*Corresponding author, email Lassaed.Hazzar@USherbrooke.ca

2016 Informa UK Limited, trading as Taylor & Francis Group


Received 21 July 2016; accepted 8 August 2016
DOI 10.1080/19386362.2016.1224030 International Journal of Geotechnical Engineering 2016 1
Hazzar et al. Lateral response of battered pile foundations

1 Typical mesh used for the 3D finite differences analyses: pile length (L)=10.0m, pile diameter (B)=1.0m and batter angle
()=25.5)

the pile shaft and sand taking into account the vertical earth layout and meshing of the FD half model used for the pile
pressure distribution. These results raise an important ques- soil system. A floating battered pile with a diameter, B and a
tion regarding to how the vertical and the lateral loads applied length, L was embedded in a soil stratum with total thickness
to the batter piles interact and how this interaction could be corresponding to Lcos+6B, where is the batter angle. The
affected by pile batter angle as well as soil types. Although, this pilesoil system was meshed with 8-node brick elements,
issue is amply discussed in the literature for both single and and the soil elements are fairly small adjacent to the pile and
grouped vertical piles (e.g. Karthigeyan et al. 2006; Achmus gradually increase in size as they move away from it. The
and Thieken 2010; Hussien et al. 2012, 2014). It represents soil element size was kept uniform at 0.5m in the vertical
a challenge for geotechnical engineers and practitioners due direction. The total mesh size was extended to a horizontal
to the complexity of the problem and the required resources distance of 16B from the centre of the pile. This distance was
and time. decided after performing a number of initial trial analyses
Aiming at filling part of this gap, we study several aspects with several horizontal distances until the displacements and
of the performance of battered piles under pure lateral loads as stresses of the pile did not change significantly with further
well as a combination of lateral and vertical loads in both sandy increase in the distance. All displacements were restrained at
and clayey soils through parametric three-dimensional (3D) the bottom of the meshes while those at the vertical external
analyses employing the finite differences (FD) code, FLAC3D faces were fully fixed in the x- and y-directions. The sym-
(Itasca 2009). The first part of the paper outlines the numerical metry face were fixed against displacement normal to the
model and shows its consistency with available experimental symmetry plane, but were free to move on the surface of the
results. plane. The top and bottom of the pile were set as displacement
and rotation free.
Finite differences modelling
Mesh details
Soil model
The behaviour of battered piles under both lateral and
combined loading conditions is investigated through three- As only the undrained behaviour of the soil was being consid-
dimensional (3D) analyses employing the finite differences ered, it was deemed sufficient to use a total stress model for the
(FD) program, FLAC3D (Itasca 2009). An advanced computer soil. This model is implemented into FLAC3D as a modification
system with parallel processor technology was used to min- of the MohrCoulomb failure criterion and requires the fol-
imise the computation time. Taking advantage of symmetry, lowing six parameters: mass density (), cohesion (c), friction
only half of the actual model was built, thus significantly angle (), dilatancy angle (), elastic bulk modulus (K) and
reducing the computational effort. Figure 1 shows the general elastic shear modulus (G).

2 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

2 Comparison between the FD predicted results and the test data of Zhang et al. (1999): a loose sand-negative angle, b loose
sand-positive angle, c dense sand-negative angle and d dense sand-positive angle

Pile model interface elements are allowed to separate if tension develops


across the interface. Once a gap is formed between the pilesoil
In this study, the pile is assumed as linear elastic material. Three interfaces, the shear and normal forces are set to zero (Itasca
parameters are required to define the pile material behaviour. 2009).
These parameters are the mass density (p), the elastic bulk In the current study, involving nonlinear analysis, high
modulus (Kp) and the elastic shear modulus (Gp). pilesoil interface stiffness is assigned to minimise the con-
tribution of pilesoil interface elements to the accumulated
pile displacements. According to the results of trial numerical
Pilesoil interface analyses conducted to identify an appropriate stiffness value, a
The pilesoil interface was simulated using the basic concept value of 108Pa/m for both kn and ks was found to be sufficient to
of the Coulomb friction model, which relates the maximum ensure that no additional deflections were attributed to the pile
allowable shear stress (friction) across an interface to the con- due to the deformation of the springs representing the interface.
tact pressure between the contacting bodies. The shear strength
was defined with zero cohesive strength and 2/3 of the friction
Numerical modelling procedures
angle for sandy soils. In the case of clayey soils, the interfaces
were assumed to have zero frictional strength and the same The model is first brought to an equilibrium stress state under
cohesive strength of the surrounding soil. Separation is able to gravitational loading before the installation of the pile. In the
cause a significant increase in displacements and therefore the next stage of analysis, the model is brought into equilibrium

 International Journal of Geotechnical Engineering2016 3


Hazzar et al. Lateral response of battered pile foundations

3 Comparison between the FD predicted results and the test data of Ranjan et al. (1980) and p-y method (Matlock, 1970): a =30,
b =0 and c =+30

after the installation of the pile. The installation is modelled by velocity at the pile top. The modelling of the pile installation
changing the properties of the pile zones from the properties process is rather complicated, so that pile is assumed to be
representing the soil material to those representing the pile in a stress-free state at the beginning of the analysis, and the
material. It is helpful to know the addresses of the grid points effect of the pile installation is not considered in the current
at the pile, to facilitate both the loading of the pile and the analyses.
monitoring of the pile response.
The ultimate vertical capacity (Vult) of the pile is evalu-
ated by applying a vertical velocity at the pile head while Table 1 Pile dimensions and material properties (CISC 2011)
the pile load and settlement are monitored. Following the
Pile details
recommendation of CGS (2013), the value of Vult is defined
in this study as the vertical load corresponding to the point Outside diameter B (mm) 406.4
Thickness T (mm) 6.3
with maximum curvature on the vertical loadvertical dis- Length L (m) 10.0
placement curve. Type of pile Aluminium
After the model is brought to equilibrium for the vertical Youngs modulus Ep (MPa) 78,000
Mass density p (kg/m3) 6260
loading, the pile top is then displaced laterally for a deflec- Poissons ratio p 0.29
tion of 0.25B. This is accomplished by applying a horizontal

4 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

Table 2 Model parameters of sandy soil used in parametric study

Mass density Shear modulus G Bulk modulus K Undrained shear Angle of friction
Soil type (kg/m3) (MPa) (MPa) strength cu (kPa) ()
Very loose 1600 4.6 10.0 0 26 (Dr=0%)
Loose 1800 7.7 16.7 30 (Dr=40%)
Dense 2000 19.2 41.7 36 (Dr=60%)
Very dense 2200 26.9 58.3 42 (Dr=87%)

Table 3 Model parameters of clayey soil used in parametric study

Mass density Shear modulus G Bulk modulus K Undrained shear Angle of friction
Clay type (kg/m3) (MPa) (MPa) strength cu (kPa) ()
Soft 1600 6.00 58.0 20 0
Medium 9.00 87.0 30
Medium 11.70 113.1 39
Stiff 19.20 185.6 64

4 Lateral loadlateral deflection curves of battered piles: a very loose sand, b loose sand, c dense sand and d very dense sand

 International Journal of Geotechnical Engineering2016 5


Hazzar et al. Lateral response of battered pile foundations

Validation of the proposed numerical


model
Before describing the numerical results on the influence of
batter angle and vertical loads on the lateral response of battered
piles, the applicability of the adopted model was verified by
predicting the pile load test data from two published cases stud-
ies. The first case corresponds to centrifuge tests on battered
piles embedded in sandy soils while the second corresponds
to laboratory tests on a model pile embedded in clayey soils.

Model pile (Zhang et al. 1999)


Lateral load tests on battered piles were conducted in a cen-
trifuge at 45g. These tests simulated prototype square piles
with 0.43-m wide and 13.7-m long founded in loose and dense
sand. Five pile inclinations were modelled: 7 and 14 at neg-
ative pile batter, plumb, and 7 and 14 at positive pile batter.
The lateral load was applied at the pile head. The free length
5 Influence of pile batter angle on lateral resistance of piles corresponding to the distance between the point of lateral load
installed in sandy soils with deferent densities application and the ground surface is 2.14m. The Youngs

6 Analyses of battered pile, with several batter angle , installed in a very dense sand: a Mohr circles of a soil element adjacent
to a Pile battered reverse and at a depth of 2m, b Mohr circles of a soil element adjacent to a Pile battered forward and at a
depth of 2m, c variation of the confining pressure along the pile stress paths of soil elements attached to the pile at different
depths and d stress paths of soil elements attached to the pile at different depths

6 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

show a fairly good comparison with the experimental results


reported by Ranjan et al. (1980). Hence, it could be concluded
that the numerical scheme adopted in the present investigation
is capable of modelling the behaviour of battered piles under
lateral loads.

Parametric studies
FLAC3D was used to perform a series of analyses on battered
piles embedded in sandy or clayey soils, and subjected to pure
lateral and combined vertical and lateral loads. The primary
objective of these analyses is to study the influence of typical
parameters on the lateral response of battered piles. Due to the
abundant number of parameters involved, this study focuses
on a selected number of parameters. These parameters include
batter angle (), vertical load, relative density of sandy soil, and
undrained shear strength, cu of clayey soil.The piles considered
in this study are made from aluminium alloy tubes with an
open end and hollow circular section. Material properties of
these piles are extracted from Handbook of Steel Construction
7 Vertical loadvertical displacement of Piles battered (CISC 2011). Table 1 presents the dimensions and the material
reverse inclined by of 12.5 and installed in sandy soils proprieties of the used piles. On the other hand, soil proper-
with deferent densities ties considered are grouped in Tables 2 and 3 for sandy and
clayey soils, respectively. For each sand density, the proposed
modulus of the model aluminium was 73.1MPa. The soil used friction angle and its corresponding relative density (Dr) are
in the study was mixed sand with average particle diameter chosen with reference to Skempton (1986) and A.P.I. (1993).
of 0.23mm. The sand layer was prepared by dry pluviation The dilatancy angle is not considered in the current analyses.
through three rectangular sieves (US standard sieve No. 14) The elastic shear modulus, G for cohesive soil was determined
which were stacked on top of the rectangular sample container. using the undrained shear strength, cu based on the correlation
Two sample densities were prepared for the tests: (1) loose sand provided by Poulos and Davis (1980):
and (2) dense sand. The dry unit weights corresponding to these
G = 300 cu (1)
relative densities were 14.05 and 14.50kN/m3, respectively.
Originally, the internal friction angles of the sands were 34.5 The response of battered piles was investigated first for sev-
and 37.1, respectively (Zhang et al. 1999). The comparisons eral batter angles () ranged from 25.5 to 25.5 (these val-
between the FD predictions and the reported data, correspond- ues correspond to the limit values where the model is stable
ing to battered piles in the loose and dense sands, are shown in the numerical analysis with respect to the adopted pile
in Fig. 2. It should be noted that the combination of loads dimensions) under pure lateral load. Then, the analyses are
and pile rigidities adopted in the previous study (Zhang et al. repeated under the combined action of vertical and lateral
1999) does not allow the piles to be deformed beyond their loads, where the value of vertical load is selected to be equal
elastic limits as it is shown in Fig. 2. The 3D numerical results to the ultimate value (Vult) of the battered pile. The combined
are fully consistent with the experimental results obtained by loads are applied in two stages. In the first stage, vertical
Zhang et al. (1999). loads were applied at pile head and then in the second stage,
lateral loads were applied while the vertical load was kept
constant. The numerical results under pure lateral loads and
Model pile (Ranjan et al. 1980) combined lateral and vertical loads on battered piles are pre-
sented and discussed separately for sandy and clayey soils in
A model aluminium pile of 9.5mm outside diameter and the following sections.
1.0-mm wall thickness, embedded 360mm in soft clay
(cu=15.2kPa) was tested by Ranjan et al. (1980). The labora-
tory test was performed on a single pile under a lateral load and Numerical modelling results and
batter angles of 30, 0 and +30. The soil Youngs modulus discussion
considered by Ranjan et al. (1980) (600kPa) was used in the
current analysis. The Poissons ratio of the clayey soil was Influence of batter angle in sandy soils
selected at 0.49 assuming an undrained response during the Figure 4(ad) shows the influence of a batter pile angle () on
load test. The lateral loadlateral deflection curves obtained the lateral response of battered piles installed in sandy soils.
from the current FD analysis are compared with the published Each plot in Fig. 4 corresponds to a different state of sand
results of Ranjan et al. (1980) and the p-y method (Matlock density including very loose, loose, dense and very dense. It
1970) as shown in Fig. 3. The experimental results agree more is appeared from Fig. 4 that the lateral response of Piles bat-
or less with the p-y method scheme with a difference increases tered forward, is not much affected compared to a vertical pile
to almost 11% at maximum lateral deflection level of 1.5mm (=0). In the case of very loose sand, the lateral response is
for the three batter angles. In spite of this, the numerical results slightly increased for of 25.5 and not significantly changed

 International Journal of Geotechnical Engineering2016 7


Hazzar et al. Lateral response of battered pile foundations

8 Lateral loadlateral deflection curves of Piles battered reverse inclined by of 12.5 and Piles battered forward inclined by
of 12.5 for the analyses with and without vertical loads: a very loose, b loose, c dense and d very dense sands

for of 12.5. In the case of very dense sand, the lateral more than 18, the lateral response increases by 6 and 4% in
response is relatively decreased for of 12.5 and 25.5. For very loose and loose sands, respectively, and decrease by 5 and
the Piles battered reverse, the lateral capacities are consid- 11% in dense and very dense sands, respectively. For negative
erably increased for of 12.5 and 25.5 in all sand state batter angle, the dependence of the battered pile response on
considered. the pile batter becomes more significant with increasing sand
The variation of the ratio of lateral capacities of battered relative density. The percentages of the increases in lateral
piles relative to vertical pile with the battered pile angle is capacities are 14, 18, 22, and reach 24% for of 12.5. For
portrayed in Fig. 5 for different states of sand densities. The of 25.5, the increases are 39, 45, 58 and reach 61% in the
general trends in Fig. 5 indicate that the lateral capacities of very loose, loose, dense and very dense sands, respectively.
battered piles depend on both the pile batter angle and the sand The reason for the considerable increase in the lateral
density. For positive batter angle, the dependence of the bat- response for Piles battered reverse and the little change for
tered pile response of the pile batter is minimal especially for Piles battered forward has been examined by plotting the
very loose and loose sands. This result confirms that proposed stress state (Mohr circle) of a soil element adjacent to the pile
by Meyerhof and Yalcin (1994). For from 0 to 18, the lateral and at a depth of 3m. The major (1) and the minor (3) prin-
response is not significantly changed relative to the correspond- cipal stresses corresponding to stress state of the soil element
ing response of vertical pile in very loose and loose sands. For at 0.25B lateral deflection of the battered piles are plotted in
dense and very dense sands, the lateral response of battered Fig. 6(a) for of 25.5, 12.5and 0 and in Fig. 6(b) for
piles however, decreases by almost 8 and 11%, respectively. For of 0, 12.5 and 25.5.

8 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

9 Influence of pile batter angle and vertical loads on the lateral resistance of battered piles in a very loose sand, b loose sand,
c dense sand and d very dense sand

Figure 6(a) illustrates that the negative batter angle considerable increase in the shear strength of the soil element.
increases the major principle stress relative to that correspond- For the case of positive batter angle, when the batter angle
ing to the vertical pile under pure lateral load. The increase in increases from 0 to 25.5, 3 slightly decreases and therefore
the major stress then increases the mobilised shear strength, fm the v0 would cause a little decrease in the shear strength of
of the soil according to: the soil element (Fig. 6(b)). Therefore, the stresses from nega-
1 3 tive inclination of pile will cause a little decrease in the lateral
fm = sin(90 + ) (2) response of the pile.
2 Considerable increase and decrease in the confining pressure
At the same lateral deflection of 0.25B, Fig. 6(a) confirms that in the soil in the vicinity of the battered pile inclined, respec-
the soil shear strength is reached for of 25.5. More lateral tively, by 12.5 and 12.5, and installed in very dense sand are
deflection of pile is needed for of 12.5 and 0 to reach shown in Fig. 6(c). This increase or decrease in the confining
failure. On the other hand, when the batter angle decreases stress of soil, then, increases or decreases the resistance of
from 0 to 25.5, 3 increases and therefore there will be a soilpile interaction.
stress increment v0 (difference between 3 corresponding to a Figure 6(d) shows the stress paths of soil elements attached
vertical pile and 3 corresponding to a battered pile with nega- to the pile installed in very dense sand and at depths of 1.0m
tive batter angle) due to the negative pile inclination. According and 2.5m for both the cases of Pile battered reverse inclined
to Zhang et al. (2002), the stress increment v0 would cause by of 12.5, Pile battered forward inclined by of 12.5

 International Journal of Geotechnical Engineering2016 9


Hazzar et al. Lateral response of battered pile foundations

and vertical pile (=0). For all considered depths, Fig. 6(d)
confirms that the soil element in the case of battered pile
inclined by of 12.5 reached the failure surface earlier than
that in the others cases corresponding of of 0 and 12.5.
Moreover, the soil element located at 1.0m reached the failure
surface before the other soil elements at deeper depths due to
the load transfer from the pile to the adjacent soil.

Influence of vertical loads in sandy soils


The ultimate vertical capacities of battered piles installed in
sandy soils with different states of density are evaluated by
applying vertical velocities at the piles heads and monitoring
the piles loads variation with their settlements as plotted in
Fig. 7 for the case of battered pile inclined by 12.5. The
value of Vult is selected as the vertical load corresponding
to the point with maximum curvature on the vertical load
vertical displacement curve as defined by CGS (2013). Figure 7
indicates that the ultimate bearing capacity of a battered pile
10 PIC vs. for several sand densities inclined by 12.5 is approximately 66, 77, 123 and 158kN
for very loose, loose, dense and very dense sand, respectively.

11 Analyses of battered pile installed in a very dense sand with and without vertical loads: a Mohr circles of a soil element adjacent
to a battered pile inclined by a =12.5and at a depth of 3m, b Mohr circles of a soil element adjacent to a battered pile
inclined by =12.5 and at a depth of 3m, c stress paths of soil elements attached to the pile inclined by =12.5 at different
depths and d stresses of a soil element adjacent to the pile inclined by =12.5 at a depth of 3m

10 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

12 Lateral loadlateral deflection curves of battered piles in clayey soil for a cu=20kPa, b cu=30kPa, c cu=39kPa and
d cu=64kPa

Figure 8 shows the influence of vertical loads on the lateral PV = v PV = 0


response of Piles battered reverse inclined by of 12.5 and PIC = 100 (3)
Piles battered forward inclined by of 12.5 installed in sandy PV = 0
soils. Each plot in Fig. 8 corresponds to a different state of sand
density including very loose, loose, dense and very dense. In where PV=v is the lateral capacity under combined loads and
the presence of vertical loads, the lateral capacities developed PV=0 is the lateral capacity under pure lateral load (without
at all deflections are less and more than the corresponding load vertical load). The variation of the PIC values with for several
developed under pure lateral load in the case of Piles battered sand densities has been represented in Fig. 10.
reverse inclined by of 12.5 and Piles battered forward It could be observed from Fig. 10 that the PIC values
inclined by of 12.5, respectively. increase with the increase in batter angle . For negative angle,
Figure 9 shows the effects of both batter angle () and ver- PIC is the same for very loose, loose and dense sands. In the
tical load (Vult) on the lateral response of battered piles in sandy case of very loose sand, PIC reaches 21.4% for of 25.5.
soils. Each plot in Fig. 9 corresponds to different states of sand For positive angle, in the case of very dense sand PIC varies
density including very loose, loose, dense and very dense. A considerably with the sand density and it reaches 45% for
quantity termed as the percentage improvement in lateral capac- of 25.5.
ity (PIC) has been defined to measure the influence of vertical The comparative results shown in Figs. 9 and 10 confirm
loads on the lateral response of battered piles. that the presence of vertical loads is as important as the batter

 International Journal of Geotechnical Engineering2016 11


Hazzar et al. Lateral response of battered pile foundations

13 Influence of pile batter angle on the lateral resistance of


battered piles in clayey soil for several cu 15 Vertical loadvertical displacement of Piles battered
reverse inclined by of 12.5 and installed in clayey soils
with deferent cu
angle in the design of battered piles foundation under lateral
loading.
The reason of the decrease in lateral capacity of Piles bat- Figure 11(c) shows the stress paths of soil elements attached
tered reverse and the increase in lateral response of Piles to the Pile battered reverse inclined by of 12.5 installed
battered forward installed in sandy soil under the action of in very dense sand and at different depths for both the cases
vertical load has been examined, firstly, by plotting the stress with (V=Vult) and without (V=0) vertical loads. For all con-
state (Mohr circle) of a soil element adjacent to the pile and sidered depths, Fig. 11(c) confirms that the soil element in the
at a depth of 3m. The major (1) and the minor (3) principal case of (V=Vult) reached the failure surface earlier than that
stresses corresponding to stress state of the soil element after in the case of (V=0). Moreover, the soil element located at
0.25B lateral deflection of the Pile battered reverse inclined 1.0m reached the failure surface before the other soil elements
by of 12.5 and Pile battered forward inclined by of at deeper depths due to the load transfer from the pile to the
12.5 installed in very dense sand (almost same results for the adjacent soil.
others sand densities) are plotted in Fig. 11(a, b), respectively, The variation of stresses (major, minor, vertical, horizontal)
for both the cases with (V=Vult) and without (V=0) vertical of a soil element adjacent to the Pile battered reverse inclined
loads. by of 12.5, installed in very dense sand and at a depth of
As expected, Fig. 11(a, b) shows that the inclusion of verti- 3m, with lateral deflections are plotted for both the cases with
cal load decreases slightly and increases, respectively, the major (V=Vult) and without (V=0) vertical loads in Fig. 11(d). Figure
principle stress relative to that corresponds to the case of a pile 11(d) indicates that the inclusion of vertical load increases the
under pure lateral load. On the other hand, the corresponding 3 soil stresses compared to the corresponding stresses in the
is very slightly increased and increased, respectively. case of pure lateral loading. Figure 11(d) shows also that the

14 Analyses of battered pile, with several batter angle , installed in a stiff clay (cu=64kPa): a Mohr circles of a soil element
adjacent to a Pile battered reverse and at a depth of 3m and b Mohr circles of a soil element adjacent to a Pile battered
forward and at a depth of 3m

12 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

16 Lateral loadlateral deflection curves of Piles battered reverse inclined by of 12.5 and Piles battered forward inclined by
of 12.5 for the analyses with and without vertical loads in clayey soil for: a cu=20kPa, b cu=30kPa, c cu=39kPa and
d cu=64kPa

orientation of the minor principle stress 3 and the major prin- moderately increased in a medium clay (cu=30 and 39kPa),
ciple stress 1 in both analyses with and without vertical loads and moderately slightly increased in a stiff clay (cu=64kPa).
are similar up to the maximum lateral deflection of 100mm. The variation of the ratio of lateral capacities of battered
piles relative to vertical pile with the battered pile angle is por-
trayed in Fig. 13 for different cu. The general trends in Fig. 13
indicate that the lateral response of battered piles depend only
Influence of batter angle in clayey soils
of the pile batter angle and not on the undrained shear strength
Figure 12 shows the influence of batter pile angle () on the cu. For the case of Piles battered forward, the dependence of
lateral response of battered piles installed in clayey soils. Each the lateral response on the pile batter is minimal especially for
plot in Fig. 12 corresponds to an shear strength cu value. The from 0 to 12.5 and for until 18 in a stiff clay (cu=64kPa).
same trend is observed as that in the case of Piles battered For of 25.5, the maximum increase in lateral response is of
reverse in sandy soils. The lateral capacities are considerably the order of 11% in very soft clay. In the case of Piles battered
increased for of 12.5 and 25.5 in for all cu considered. In reverse, the dependence of the lateral response of the pile bat-
the case of Piles battered forward, the lateral response is not ter is almost the same in all cu. For of 25.5, the maximum
significantly changed for of 12.5. For of 25.5, the lateral increase in lateral response is of the order of 25% in stiff clay
response is considerably increased in a soft clay (cu=20kPa), (cu=64kPa).

 International Journal of Geotechnical Engineering2016 13


Hazzar et al. Lateral response of battered pile foundations

17 Influence of pile batter angle and vertical load on the lateral resistance of battered piles in clayey soil for: a cu=20kPa,
b cu=30kPa, c cu=39kPa and d cu=64kPa

Similar to the case of sandy soil, the mechanism of the =25.5 and 12.5 in the case of Piles battered reverse
considerable increases in the lateral response of Piles battered and =12.5 and 25.5 in the case of Piles battered forward.
reverse and the little increases in the lateral response of Piles Thus, the batter angle increases the lateral resistance of soil and
battered forward installed in clayey soil has been examined subsequently will lead to development of power resistance to
by plotting the stress state (Mohr circle) of a soil element adja- lateral pile deformation. These observations illustrate well the
cent to the battered pile and at a depth of 3m. The ultimate results given in Fig. 12.
shear stress ult, corresponding to the failure, is calculated by
the following equation: Influence of vertical loads in clayey soils
1
ult = 1 = cu (4) Similar to clayey soil cases, the ultimate vertical capacities of
2 piles installed in clayey soils with different cu are evaluated by
The 1 and 3 corresponding to stress state of the soil ele- applying vertical velocities at the piles heads and monitoring
ment before the application of lateral load on a Pile battered the piles loads variation with their settlements as plotted in
reverse and a Pile battered forward installed in a stiff clay Fig.15. Figure 15 shows that Vult of the battered pile inclined
(cu=64kPa) are plotted in Fig. 14(a, b), respectively, for sev- by 12.5 and installed in clayey soil is approximately 84, 144,
eral . It is clear that the Mohr circles corresponding to vertical 185 and 329 kN in soft (cu=20kPa), medium (cu=30kPa),
pile (=0) have larger radius than those corresponding to medium (cu=39kPa) and stiff (cu=64kPa) clays, respectively.

14 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

Figure 16 shows the influence of vertical loads on the lateral


response of Piles battered reverse inclined by of 12.5
and Piles battered forward inclined by of 12.5 installed
in clayey soils. Each plot in Fig. 16 corresponds to a different
state of clay rigidity including soft with cu of 20kPa, medium
with cu of 30kPa and cu of 39kPa, and stiff with cu of 64kPa.
In the presence of vertical loads, the lateral response developed
at all deflections are less and more than the corresponding load
developed under pure lateral load in the case of Piles battered
reverse inclined by of 12.5 and Piles battered forward
inclined by of 12.5, respectively.
Figure 17 shows the effects of both batter angle () and ver-
tical load (Vult) on the lateral response of battered piles in clayey
soils. Each plot in Fig. 17 corresponds to different rigidity of
clay including soft, medium and stiff. As the case for sandy soil,
Fig. 18 shows the variation of the percentage improvement in
lateral capacity (PIC) according to the batter angle value for
several cu of clay. It could be observed from Fig. 18 that the
PIC values increase with the increase in batter angle . While
the increase in cu doesnt appear to have significant effect. The
18 PIC versus for several cu of clay
maximum PIC reaches 15.1% for of 22.5 in the case of a

19 Analyses of battered pile installed in a rigid clay (cu=64kPa) with and without vertical loads: a Mohr circles of a soil element
adjacent to a battered pile inclined by a =12.5and at a depth of 3m, b Mohr circles of a soil element adjacent to a battered
pile inclined by =12.5 and at a depth of 3m, c stress paths of soil elements attached to the pile inclined by =12.5 at
different depths and d stresses of a soil element adjacent to the pile inclined by =12.5 at a depth of 3m

 International Journal of Geotechnical Engineering2016 15


Hazzar et al. Lateral response of battered pile foundations

medium clay with cu of 39kPa, and 9.4% for of 22.5 in the For a range of more than 18, the lateral capacities
case of a stiff clay with cu of 64kPa. increases by 6 and 4% in very loose and loose sands,
The comparative results shown in Figs. 17 and 18 confirm respectively, and decrease by 5 and 11% in dense and
that the presence of vertical loads is as important as the batter very dense sands, respectively. In the case of Piles bat-
angle and the undrained shear strength of clay in the design of tered reverse, the lateral capacities are considerably
battered piles foundation under lateral loading. increased with the increase in both and sand density.
The reason of the decrease in the lateral response of Pile The percentages of these increases reach 39, 45, 58 and
battered reverse and the increase in the lateral response of 61% for of 25.5 in very loose, loose, dense and very
Pile battered forward in clayey soil and under the action of dense sands, respectively.
vertical loads has been also examined through variations of (2)The influence of vertical load (corresponding to the
Mohr circles of a soil element adjacent to the pile and at a depth ultimate vertical load) on lateral response of battered
of 3m. The 1 and 3 corresponding to stress state of the soil piles in sandy soils is very important. The percentage
element after the application of vertical load and before the improvement in lateral capacity (PIC) increases with
application of lateral load on a Pile battered reverse inclined the increase in value. For positive batter angles, PIC
by of 12.5 and a Pile battered forward inclined by of varies considerably with the sand density and reaches
12.5 installed in clay are plotted in Fig. 19(a, b), respectively, 45% for of 25.5 in the case of very dense sand. For
for both the cases with (V=Vult) and without (V=0) and for cu negative angles, PIC is almost the same for very loose,
of 64kPa. When Vult is applied to the Pile battered reverse, loose and dense sands and reaches 21.4% for of
the Mohr circle has larger radius than this corresponding to 25.5 in the case of very loose sand.
(V=0). Thus, the presence of the vertical loads decreases the (3)In the case of clayey soils, the lateral response of battered
lateral resistance of soil and subsequently will lead to devel- piles depends only on the batter angle and does not vary
opment of lower resistance to lateral pile deformation. For the with the undrained shear strength cu of clay. For less
case of Pile battered forward, when Vult is applied, the Mohr 12.5, the lateral capacities of Piles battered forward are
circle has smaller radius than this corresponding to (V=0). not changed for all cu considered. But, when is more
Thus, the presence of the vertical load increases the lateral 12.5, these lateral capacities are moderately increased.
resistance of soil and subsequently will lead to development of The maximum increase of these lateral capacities is of
power resistance to lateral pile deformation. Figure 19(c) shows the order of 11% in the case of a very soft clay for
the stress paths of soil elements attached to the Pile battered of 25.5. In the case of Piles battered reverse, the lat-
reverse inclined by of 12.5 in stiff clay (cu=64kPa) at eral capacities are increased and the maximum increase
different depths for both the cases with (V=Vult) and without reaches 25% in the case of a stiff clay for of 25.5.
(V=0) vertical loads. For all considered depths, the soil ele- (4)The effects of vertical load on the lateral response of
ment in the case of (V=Vult) reached the surface failure earlier battered piles in clayey soils is moderately important.
than the soil element in the case of (V=0). The variation of The PIC values increase with the increase in and the
stresses (major, minor, vertical, horizontal) of a soil element increase in cu doesnt appear to have significant effect
adjacent to the Pile battered reverse inclined by of 12.5 on the PIC value.
and installed in stiff clay (cu=64kPa) at a depth of 3m, with It is fair to mention that the current parametric analyses
lateral deflections are plotted for both the cases with (V=Vult) dont account for the pile flexibility and installation method (i.e.
and without (V=0) vertical loads in Fig. 19(d). Figure 19(d) driving or drilling of the pile). Consequently, the above conclu-
indicates that the inclusion of vertical load decreases the soil sions may not be applied for soilpile systems with character-
stresses compared to the corresponding stresses in the case of istics substantially different from those adopted in this study.
pure lateral loading.
References
Conclusions A.P.I. American Petroleum Institute. 1993. Planning, designing and constructing
The influence of the vertical loads on the behaviour of laterally fixed offshore platforms, RP2A-LRFD, Section G, 6477.
loaded battered piles in sandy and clayey soils was investigated Achmus, M. and Thieken, K. 2010. On the behavior of piles in non-cohesive
soil under combined horizontal and vertical loading. Acta Geotechnica,
by means of numerical modelling using the computer program 5, (3), 199210.
FLAC3D. The verified numerical model was used to perform a Alizadeh, M. and Davisson, M. T. 1970. Lateral load test on piles-Arkansas
parametric study with different batter angles and soil rigidities river project. Journal of the Soil Mechanics and Foundations Division ASCE,
to evaluate the lateral response of battered piles subjected to 96, (5), 15831603.
Awad, A. and Ayoub, A. 1976. Ultimate uplift capacity of vertical and inclined
lateral as well as combined loading. Based on the results of piles in cohesionless soil, Proceedings of the 5th Conference on Soil
this parametric study, the following conclusions can be made. Mechanics and Foundation Engineering. Budapest, Hungary, 1, 221227.
Canadian Geotechnical Society. 2013. Canadian Foundation Engineering
(1)The lateral response of the battered piles in sandy soils
Manual. 4th edn, Richmond, Canadian Geotechnical Society.
under lateral loads is influenced by both pile batter Canadian Institute of Steel Construction. 2011. Handbook of Steel Construction-
angle, and sand density. For a range of from 0 to Hollow structural sections G40.20-Round. 10th edn. Markham, Canadian
18, the lateral response of Piles battered forward is Institute of Steel Construction.
not significantly changed relative to the corresponding Hanna, A. M. and Afram, A. 1986. Pull-out capacity of single batter piles in
sand. Canadian Geotechnical Journal, 23, (3), 387392.
capacities of vertical piles in very loose and loose sands. Hussien, M. N., Tobita, T., Iai, S. and Karray, M. 2014. On the influence of
However, for dense and very dense sands, these lateral vertical loads on the lateral response of pile foundation. Computers and
capacities decreases by almost 8 and 11%, respectively. Geotechnics, 55, (1), 392403.

16 International Journal of Geotechnical Engineering2016


Hazzar et al. Lateral response of battered pile foundations

Hussien, M. N., Tobita, T., Iai, S. and Rollins, K. M. 2012. Vertical loads Murthy, V.N.S. 1965. Behaviour of battered piles embedded in sand subjected
effect on the lateral pile group resistance in sand. Geomechanics and to lateral loads, PhD thesis, Indian Institute of Technology, Kharagpur, India.
Geoengineering, 7, (4), 263282. Poulos, H. and Davis, E. H. 1980. Pile foundation analysis and design,
Itasca. 2009. Users and theory manuals of FLAC3D: fast Lagrangian analysis Singapore, Wiley.
of continua in 3D, version 4. Minneapolis, MN, Itasca Consulting Group Inc. Ranjan, G., Ramasamy, G. and Tyagi, R. P. 1980. Lateral response of batter
Karthigeyan, S., Ramakrishna, V. V. G. S. T. and Rajagopal, K. 2006. Influence piles and pile bents in clay. Indian Geotechnical Journal, 10, (2), 135142.
of vertical load on the lateral response of piles in sand. Computers and Skempton, A. W. 1986. Standard penetration test procedures and the effects
Geotechnics, 33, (2), 121131. in sands of overburden pressure, relative density, particle size, ageing and
Lu, S. S. 1981. Design load of bored pile laterally loaded, Proceedings of the l0th overconsolidation. Gotechnique, 36, (3), 425447.
International Conference on Soil Mechanics and Foundation Engineering, Veeresh, C. 1996. Behavior of batter piles in marine clays, PhD thesis, Indian
Balkema. Rotterdam, The Netherlands, 767777. Institute of Technology, Madras, India.
Matlock, H. 1970. Correlations for design of laterally loaded piles in soft clay, Zhang, L. M., McVay, M. C., Han, S. J. and Lai, P. W. 2002. Effects of dead
2nd Offshore Technology Conference, Texas, Houston, 577594. loads on the lateral response of battered pile groups. Canadian Geotechnical
Meyerhof, G. G. and Ranjan, G. 1973. The bearing capacity of rigid piles under Journal, 39, (3), 561575.
inclined loads in sand. II: batter piles. Canadian Geotechnical Journal, Zhang, L. M., McVay, M. C. and Lai, P. W. 1999. Centrifuge modelling of
10, 7185. laterally loaded single battered piles in sands. Canadian Geotechnical
Meyerhof, G. G. and Yalcin, A. S. 1994. Bearing capacity of flexible batter piles Journal, 36, (6), 10741084.
under eccentric and inclined loads in layered soil. Canadian Geotechnical
Journal, 31, 583590.

 International Journal of Geotechnical Engineering2016 17