Beruflich Dokumente
Kultur Dokumente
DOI 10.1007/s10706-006-9110-7
ORIGINAL PAPER
Received: 11 June 2005 / Accepted: 20 September 2006 / Published online: 27 October 2006
Springer Science+Business Media B.V. 2006
Abstract Lateral load-deflection behaviour of parametric study undertaken on the basis of the
single piles is often analysed in practice on the proposed P–Y curves showed the significant influ-
basis of methods of load-transfer P–Y curves. ence of the lateral pile/soil stiffness on the non-
The paper is aimed at presenting the results of linear load-deflection response.
the interpretation of five full-scale horizontal
loading tests of single instrumented piles in two Keywords Lateral loading test Æ Lateral reaction
sandy soils, in order to define the parameters of modulus Æ P–Y curves Æ Pressuremeter test Æ Sand Æ
P–Y curves, namely the initial lateral reaction Single pile
modulus and the lateral soil resistance, in correla-
tion with the pressuremeter test parameters. P–Y List of symbols and units
curve parameters were found varying as a power of B diameter or frontal width of the pile (m)
lateral pile/soil stiffness, on the basis of which D embedded length of the pile (m)
hyperbolic P–Y curves in sand were proposed. The De effective pile length (m)
predictive capabilities of the proposed P–Y curves E elastic soil modulus (MPa)
were assessed by predicting the soil/pile response e excentricity of lateral load (m)
in full-scale tests as well as in centrifuge tests and a Ec characteristic soil modulus (MPa)
very good agreement was found between the Em first load pressuremeter modulus (MPa)
computed deflections and bending moments, and Er reload pressuremeter modulus (MPa)
the measured ones. Small-sized database of full- Eti initial lateral reaction modulus (MPa)
scale pile loading tests in sand was built and a EpIp flexural pile stiffness (MN m2)
comparative study of some commonly used P–Y F tangential lateral reaction (kN/m)
curve methods was undertaken. Moreover, it was Fl limit tangential lateral reaction or
shown that the load-deflection curves of these test tangential lateral resistance (kN/m)
piles may be normalised in a practical form for an Gr pressuremeter shear modulus (Gr = Er/
approximate evaluation of pile deflection in a [2(1 + m)]) (MPa)
preliminary stage of pile design. At last, a H lateral load applied on the pile top (kN)
Id density index (%)
K pile/soil compressibility
A. Bouafia (&)
Kr lateral pile/soil stiffness
Department of civil engineering, University of Blida,
P.O. Box 270, R. P Blida 09000, Algeria L tangential dimension of the pile section
e-mail: bouafia1@yahoo.fr (parallel to H) (m)
123
284 Geotech Geol Eng (2007) 25:283–301
L0 transfer length or elastic length (m) foundations were besides horizontally loaded,
M bending moment at a given depth (kN m) inclined piles, often difficult to achieve were to
M0 bending moment applied to the pile top be added. Due to the progress done in the
(kN m) knowledge of piles foundation behaviour, it is
NH rate of increase of Eti with depth in nowadays recognised that vertical piles can sus-
Gibson’s soil (MPa/m) tain horizontal loads. Earth pressures on a bridge
Nspt N value of the SPT (blow counts/30 cm) abutment piles, lateral displacement of soft clayey
P lateral soil reaction at a given depth layer underlying an access embankment to a
(kN/m) motorway and wind pressures on slender struc-
Pu lateral soil resistance or limit lateral tures built on piles are usual examples of hori-
reaction (kN/m) zontal loading of piles. Behaviour of piles under
pf pressuremeter creep pressure (kPa) horizontal loads is a complex soil/pile interaction
pl limit pressuremeter pressure (kPa) problem because of the tridimensional nature of
p* net limit pressuremeter pressure (kPa) the phenomenon and its dependence on a multi-
Ple net equivalent limit pressuremeter tude of key parameters. This fertile domain of
pressure (kPa) research was investigated since more than a half
p0 at-rest lateral earth pressure (kPa) century.
Q frontal lateral reaction (kN/m) Geotechnical literature contains a wealth of
qc cone penetration resistance (MPa) methods of analysis mainly based on elasticity,
qs limit skin friction along the pile shaft finite/boundary elements or on subgrade reaction
(kPa) theory. However, It should be emphasised that
R least-squares regression coefficient (%) the theoretical approaches offer simplistic
R0 initial radius of pressuremeter borehole schemes of soil/pile response and therefore do
(m) not take into consideration many pile/soil inter-
DR increase in PMT borehole radius (mm) action parameters such as the pile installation, the
Sf , S t shape factors soil/pile interface roughness and the soil com-
Y lateral displacement or deflection at a pressibility. Furthermore, some particular aspects
given depth (mm) of the problem of laterally loaded piles such as
Y0 pile deflection at ground level (mm) the proximity of a slope, the group effects, and
Yref. reference deflection or threshold of lateral piles undergoing lateral soil movement are diffi-
soil resistance Pu(mm) cult to be analysed by theoretical methods.
z depth with respect to the ground level (m) Experimental research may then be considered
zc critical depth (m) as the most adapted way to investigate such a
k rate of linear increase of Em with depth in problem. The last four decades were marked by a
Gibson’s soil (MPa/m) considerable progress in the understanding of the
l rate of linear increase of Pl with depth in response of a pile to bending forces by means of
Gibson’s soil (kPa/m) several experimental studies in full-scale as well
g lateral resistance factor as in centrifuge.
m Poisson’s ratio Prior to the development of numerical meth-
w ratio Eti to Em ods in geotechnical engineering, piles were usu-
n ratio Pu to PL B ally designed by evaluating the deflections under
working loads on the basis of small displacement
methods such as the elasticity. Subgrade reaction
theory was also used for the linear analysis of pile
deflection by modelling the pile as a beam on
1 Introduction elastic foundations. These approaches were
adapted for simple pile/soil configuration and do
Pile foundations were initially designed in order not account for the soil properties heterogeneity
to transmit vertical loads to the soil. When these and the non-linear lateral response of the pile/soil
123
Geotech Geol Eng (2007) 25:283–301 285
system. Moreover, foundations of some structures Kr. Based on these relationships, hyperbolic
working under severe lateral loading conditions functions were proposed to describe P–Y curves.
are designed on the basis of limit equilibrium Validation process was undertaken by comput-
methods. These latter ones are based on approx- ing the tests piles used to derive such a method as
imate mechanisms of soil resistance derived from well as other test piles in sandy soils. Comparative
the lateral earth pressures theory (Bouafia 1990, study showed the good prediction capability of
1998; Bouafia et al. 1991). the proposed soil/pile stiffness dependant P–Y
It is nowadays recognised that the design curve methods compared to the current
methods based on P–Y curves are the most approaches based on the PMT test.
reliable to the analysis of the behaviour of
laterally loaded single piles with possibility of
taking account of the non-homogeneity of soil
2 Brief review of the methods of construction
properties as well as of the material non linearity
of P–Y curves
in lateral pile/soil response. Soil/pile interface is
modelled by infinity of non-linear springs in
To the knowledge of the author, the first study on
which the soil reaction P at a given depth is
the basis of P–Y curves was due to Reese and
undertaken by the spring for a lateral pile
Matlock (1977) by introducing the concept of the
displacement Y.
lateral reaction modulus previously defined by
Full-scale tests on instrumented piles are often
Winkler (1867). The first generation of P–Y
used to investigate the soil/pile response in the
curves was bilinear describing an elastic plastic
light of load-transfer theory. P–Y curves are
behaviour at the pile/soil interface.
derived from bending moment profiles measured
The in-situ tests such as the PMT become
by strain gauges along the pile. However, a few
usual tools for pile foundations analysis and
full-scale tests on instrumented piles in sand
design. The PMT test provides an experimental
were reported in the literature with successful
stress–strain curve describing the borehole
derivation of P–Y curves from double differen-
response under radial loads. Some similitude
tiation and integration of the bending moment
exists between the expansion of the PMT beor-
profile. The main difficulty in deriving these
ehole and the mobilisation of the frontal lateral
curves is due to the high sensitivity of the lateral
reaction of the soil around a pile (Ménard et al.
soil reaction P to the experimental conditions as
1969).
well as to the method of fitting and differenti-
Geotechnical literature contains a diversity of
ation of bending moments (Bouafia and Garnier
methods for deriving P–Y curves from PMT
1991).
parameters, namely the PMT deformation mod-
This paper is aimed at presenting the results of
ulus Em and the limit PMT pressure pl. For
an extensive analysis of full-scale horizontal piles
brevity, only the commonly used methods will be
loading tests in quite homogeneous sandy soils.
presented hereafter.
Test piles were well instrumented, and P–Y
curves were derived from the interpretation of
bending moment distribution along the experi- 2.1 Method of Ménard et al. (1969)
mental pile. The experimental results presented
herein are part of an important research pro- This method was initially suggested by Ménard
gramme carried out by the LCPC (Laboratoire et al. (1969), and then improved by Gambin
Central des Ponts & Chaussées, France) during (1979). As illustrated in Fig. 1, the curve 1 is tri-
more than three decades. linear shaped. The first portion has a slope equal
It was shown the existence of fundamental to the initial lateral reaction modulus Eti, the
relationships between the P and Y curves param- second one has a slope equal to the Eti/2 and the
eters namely the lateral soil modulus and the third one corresponds to the lateral soil resistance
lateral soil resistance, the parameters of pressure- taken equal to net limit pressure multiplied by the
meter test (PMT) and the lateral pile/soil stiffness diameter (or the frontal width) B.
123
286 Geotech Geol Eng (2007) 25:283–301
*
Pu=Pl .B.ξ(Kr) to neglect the effect of the slenderness ratio D/B
3 of the pile, D being the embedded length of the
*
Pl .B 1 pile.
P (kN/m)
Eti 18B
¼ for B[B0 ð2Þ 2.2 Method of the French code Fascicule-62
Em 4B0 ð2:65 BB Þa þ 3Ba
0 (MELT 1993)
a is a rheological factor called ‘‘coefficient of soil The previous method was integrated in the French
structure’’ depending on the nature of the soil and geotechnical code with reduction of the lateral soil
its compressibility. It is equal to 1/3 for loose and resistance to the net creep pressure pf* multiplied
medium dense sands and 1/2 for very dense sands. by B, as illustrated by curve 2 in Fig. 1. This
For small diameter piles (B £ 0.60 m) ratio Eti/ adaptation was dictated by the necessity to obtain
Em ranges between 2.24 and 2.75. conservative prediction of the pile response at
Pressuremeter parameters defining the P–Y large deflections (Baguelin et al. 1978).
curves above a critical depth zc, should be For non-circular pile section, in addition to the
reduced to take into consideration a reduction lateral reaction defined by curve 2 in Fig. 1,
in soil resistance due to soil heave and a probable tangential lateral reaction F is mobilised along the
reduction in soil confinement (Baguelin et al. tangential sides. Lateral tangential F–Y curve is
1978; Briaud 1986). According to Ménard (1971), defined as a bilinear curve. The first linear portion
the critical depth zc is equal to 4 diameters in has a slope equal to that of P–Y curve and the
granular soils and to 2 diameters in cohesive soils second one is horizontal and represents the
(Frank and Jézéquel 1989). tangential lateral resistance Fl given by
It should be emphasised that the model of
lateral reaction proposed by Ménard (1971) is Fl ¼ 2qs ðL BÞ ð3Þ
limited to a rigid pile section and therefore
neglects the effects of the pile flexural rigidity. qs is the limit skin friction equal to that mobilised
Moreover, analogy assumed by Ménard (1971) under vertical loads and L is the tangential
between the pile and an infinitely long beam leads dimension of the pile section. The overall P–Y
123
Geotech Geol Eng (2007) 25:283–301 287
curve is the superposition of the two lateral The shape factor St is equal to 2 for square piles
reaction curves. and to 1 for circular pile section. Limit skin
friction qs slightly differs from the one mobilised
2.3 Method of Dunand (1981) along the pile shaft under vertical loads and then
may be evaluated with usual bearing capacity
This method is based on a bi-linear P–Y curve as formulae (Smith 1987).
illustrated by curve 2 in Fig. 1. Lateral reaction According to the authors, the assessment of
modulus was correlated to Em on the basis of an this method with respect to the experimental
elastic method whereas the limit lateral reaction evidence of 27 pile loading tests carried out in a
is equal to plB. The concept of critical depth is variety of piles and soils showed a good predictive
introduced as in the previous methods. According capability of pile deflections (Briaud 1986).
to the author, this method is recommended to the
design of drilled piers supporting transmission 2.5 Method of Baguelin et al. (1978)
line structures.
In this method, based on the self-boring pres-
suremeter test (SBPMT), the total P–Y curve is
2.4 Method of Briaud et al. (1985) constructed point by point from the experimental
PMT expansion curve as follows:
The total lateral reaction P to the deflection Y at
a given depth is the sum of the frontal reaction Q F ¼ g p B ð7Þ
and the tangential reaction F. As a result, the P–Y
is the addition of the Q–Y curve and the F–Y BDV
Y¼ ð8Þ
curve. Carayannacon et al. (1979) showed by 4V0
finite element analysis that the contribution of
the tangential reaction increases with slenderness V0 and DV are respectively the initial PMT
of the pile section. borehole volume and the increase in borehole
The main assumption in this method is that volume under the net pressure p* g is called
radial displacements of a PMT borehole and the lateral resistance factor taking into account the
pile deflections are homothetic. Q–Y is directly surface effect and varies form 0.33 to 3 (Baguelin
built from the expansion curve of the PMT test as 1982).
follows:
2.6 Method of Robertson et al. (1984)
Q ¼ Sf p B ð4Þ
P–Y curve is constructed for a bored pile from a
BDR prebored PMT or a self-boring PMT, and for a
Y¼ ð5Þ
2R0 driven pile from a pushed-in PMT test. Formulae
5 and 7 should be used with factor g equal for
The shape factor Sf is equal to 1 for square sandy soils to 0 at surface and increasing linearly
piles and to p/4 for circular section. R0 and DR are with depth to 1.5 at the critical depth and below.
respectively the initial PMT borehole radius and The critical depth was estimated to 4 diameters
the increase in borehole radius under the net (Robertson et al. 1985).
pressure p*. Atukoralla and Byrne (1984) analysed by finite
F–Y curves have a bilinear shape composed of element modelling the lateral displacements of a
an initial portion with a slope equal to 2Gr, Gr rigid disk within an elastic plastic material as well
being the PMT shear modulus, and a horizontal as those due the cylindrical cavity expansion
asymptote equal to the limit frontal reaction Fl as within the same material. It was shown that the
follows: ratio of lateral pressures surrounding the disk to
those around the PMT cavity varies between 1.4
F l ¼ St qs B ð6Þ and 1.7 with an average value of 1.50, which is in
123
288 Geotech Geol Eng (2007) 25:283–301
accordance with the factor g of this method. 8% corresponding to a saturation degree of 31%.
However, results of this study do not account for It was possible to recover some samples with a
the tridimensional response of the pile under 150 mm diameter auger sampler up to 4.0 m of
lateral loads. depth. The density index Id measured according
As summarised in Table 1, the ratio of lateral to ASTM standard is 66%. Profiles of PMT, CPT
soil resistance Pu to PlB proposed by the methods and DPT tests are illustrated in Fig. 3.
mentioned above ranges in a wide margin betw-
een 0.3 and 3, which shows some uncertainty in 3.2 Test piles
predicting the soil resistance. As an alternative,
the Experimental analysis of instrumented test Test piles are steel pipes instrumented by strain
piles will be used in the next section to evaluate gauges distributed by pairs along two diametri-
the lateral soil resistance. cally opposite axes. Table 2 summarises the
main geometrical and mechanical characteristics
of the piles. Three tubes, noted T5, T10 and T15
3 Presentation of full-scale tests in sand were tested in site S1 and two piles P1 and P2
tested in site S2. The slenderness ratio (embed-
3.1 Sites and geotechnical conditions ded length D/diameter B) varies between 5.5
and 15.3.
The first site, noted S1, is located in Châtenay-sur- Piles in site S1 are externally instrumented by
Seine, 70 km south east of Paris (France). A big seven pairs of strain gauges irregularly distributed
pit whose volume is 424 m3 was previously dug to along two diametrically opposite axes and pro-
a depth of 3.20 m in a chalky soil. It was tected by an adhesive papers of aluminium.
waterproofed by plastic sheets, and then filled in Figure 4 illustrates a general view of piles in site
by Fontainebleau sand into two medium dense S1 with the scheme of loading device. Each pile in
layers. The underlying layer is 1.40 m thick with a this site was connected at its tip by a 90 jacking
density index Id = 37% whereas the upper layer cone in order to facilitate the jacking process into
has a thickness of 1.80 m and Id = 57%. Fon- soil. As shown in Fig. 4, the cone has same
tainebleau sand is poorly graded sand. In-situ diameter as that of the pile. Each pile was jacked
tests, notably PMT (Ménard pre-bored pressure- by means of a hydraulic jack in contact with a
meter test), CPT (static cone resistance test) and reaction beam (Canepa et al. 1987).
DPT (dynamic penetration test) were carried out Piles P1 and P2 in site S2 are instrumented by
and typical profiles are shown in Fig. 2. 20 and 22 pairs of strain gauges, respectively.
The second site, noted S2, is located in Le-Rheu, These latter are externally placed along the pile
5 km south west of Rennes (France). The soil is P1 and protected by steel valley for each axis. For
composed of reddish poorly graded clean sand of pile P2, they are internally stuck along two axes
marine origin from the Pliocene era. Ground along the pile. For both the piles strain gauges are
water table was found at 10 m of depth. The sand regularly spaced of 25 cm and the first one
above water table has average water content of corresponds to the ground surface. Each pile
was placed into a borehole previously made by a
helical drilling engine 6 m high. Some irregularity
Table 1 Comparison of theoretical ratios Pu/plB in diameter of borehole within 5 cm was noticed.
Method Pu/plB Remarks It was likely due to a default of verticality of
drilling axis (Jézéquel 1988). Each pile in site S1
Ménard et al. (1969) 1.00 was filled in with bentonite-cement grout. The
Fascicule-62 0.50 Usual correlation
pf pl/2 diameter of pile was directly measured at surface
Dunand (1981) 1.00 as well as estimated from the volumes of steel and
Briaud et al. (1982) 0.83 Bored pile in sand bentonite-cement. Uniaxial compression tests of
Baguelin et al. (1977) 0.3–3.0 bentonite samples have given a Young’s modulus
Robertson et al. (1984, 1985) 1.50 Beyond 4B
of 3,500 MPa at 28 days.
123
Geotech Geol Eng (2007) 25:283–301 289
123
290 Geotech Geol Eng (2007) 25:283–301
bending moments curve for a given load at pile were used to fit the bending moment distribution.
top. Two successive integrations of this curve lead The fitting function was chosen according to the
to easily determine lateral displacement Y along criterion of static equilibrium of the test pile under
the pile. Moreover, two successive differentia- lateral reaction profile P(z) and the loads on the
tions of this curve allow the determination of pile top within a given tolerance (Bouafia and
horizontal soil reaction P and then to define P–Y
curve at any depth.
Since the soil reaction geometrically represents
the curvature of bending moment distribution, it is
therefore very sensitive to any fluctuation of
bending moment at a given depth and strongly
depends on the choice of the fitting curve of
bending moment (Bouafia 1990; King 1994).
Quintic spline functions or polynomial functions
123
Geotech Geol Eng (2007) 25:283–301 291
Garnier 1991). This criterion was subsequently (Reese 1971; Garassino 1976; Georgiadis et al.
adopted in other studies in LCPC (Mezazigh 1995; 1992). Experimental P–Y curves were fitted by
Remaud 1999). the following hyperbolic function:
123
292 Geotech Geol Eng (2007) 25:283–301
Fig. 6 Typical P–Y Pile P1 site Le-Rheu B=0.5 m D/B=10 Ep.Ip=56.37 MN.m
2
Fig. 7 Comparison of
40
computed and measured Pile T15 measured
deflections in site S1 35 computed from
P-Y curves
Lateral load (kN)
30
25
Pile T10
20
15
10
Pile T5
5
0
0 20 40 60 80 100 120 140 160 180
Top deflection (mm)
123
Geotech Geol Eng (2007) 25:283–301 293
pffiffiffiffiffiffi
Pu ¼ 3PL B Kr ð16Þ
Eti 5 1
ðD=BÞ2 pffiffiffiffi ð17Þ
Em 4 K
pffiffiffiffi
Pu 2 K
¼ ð18Þ
pL B 3 ðD=BÞ2
b * b
Eti/Em=a.(Kr) 0,1 Pu/(PL.B)=a.(Kr)
a=0.28
b=-0.55 a=3.0
R=94 % b=0.50
R=94%
1 0,01
0,01 0,1 1E-3 0,01 0,1 1
Kr Kr
Fig. 9 Variation of the ratio Eti/Em with lateral pile/soil Fig. 10 Variation of the ratio Pu/PL B with lateral pile/soil
stiffness Kr stiffness Kr
123
294 Geotech Geol Eng (2007) 25:283–301
123
Geotech Geol Eng (2007) 25:283–301 295
Fontainebleau – 16.7 473.60 Driven 34.1 3.0 · 10–3 7.33 (Remaud 1999)
Le-Rheu ENSM-1 15.6 44.70 Bored 7.60 9.4 · 10–3 3.80 (Bouafia 1987)
Le-Rheu ENSM-2 15.6 44.70 Bored 10.3 6.9 · 10–3 4.16 (Bouafia 1987)
Le-Rheu P1–2 10.0 56.60 Bored 3.60 2.5 · 10–2 2.15 (Bouafia 1990)
Le-Rheu P1–4 10.0 56.60 Bored 11.6 7.8 · 10–3 4.20 (Bouafia 1990)
Le-Rheu P1–t 10.0 56.60 Bored 18.7 4.8 · 10–3 4.30 (Bouafia 1990)
uniformly graded. Piles TG and TD are installed of predictions are encouraging seeing the multi-
and loaded as in the above site. tude of approximations made during the process
The Roosevelt bridge site is composed of loose of definition of this method.
layer of sand thick of 4 m, overlying a thick layer Lateral response of test pile T3 in Lock & dam
of very dense partially cemented sand. The site 26 and piles 2 and 16 in the Arkansas River was
with submerged by water up to 2 m above the predicted. As illustrated by Fig. 13, the ratio
ground level. Square prestressed concrete pile Ypred.
0 /Ymeas.
0 was found fluctuating around 1.33
was driven and tested up to cracking under a load within an interval 094–2.80 and a coefficient of
of 200 kN and concrete failure occurred under a variation of 36%.
load of 320 kN. Sandy materials of sites S1 and S2 were used in
Pile T3 was tested in lock & Dam site 7 years the LCPC centrifuge to study the lateral behaviour
prior to tests on piles 3–12 and 3–13, PMT data of centrifuged models in sand within the scope of an
were not available. Prediction of the pile T3 was important programme of research undertaken by
made with the PMT data of piles 3–12 and 3–13. the LCPC since two decades. Reduction scales of
In the Arkansas site, the soil is a saturated SP/ piles were 1/40 and 1/18 for models in Fontaine-
SM sand and only the SPT test was carried out. bleau and Le-Rheu sands respectively. Character-
PMT data were estimated by usual correlation istics summarised in Table 4 correspond to the
with the SPT. prototype ones. Sandy mass was characterised by
It should be emphasised that the reliability of cone penetration tests (CPT) carried out by min-
the predictions of piles T3 in Lock & Dam and iature cone during the centrifugation. CPT tests in
piles 2 and 16 in Arkansas site will decrease centrifuge were used to estimate the PMT data by
because of the scatter in the estimation of the adopting the same correlation CPT/PMT found
PMT data for pile T3 or in the correlation with the in-situ. This assumption leads to a rough estimation
SPT test for piles 2 and 16. In this regard,
predictions of these piles will interpreted sepa-
rately. 25
1 : Loose sand Em/PL*=5
For each pile, lateral pile/soil stiffness was 1
2 : Medium dense sand Em/PL*=10
evaluated and hyperbolic P–Y curves according 20
3 : Very dense sand Em/PL*=20
to the Eqs. 14 and 16 were defined. In most of
cases, piles were sufficiently long to be considered
Yref./B (%)
15
123
296 Geotech Geol Eng (2007) 25:283–301
Fig. 12 Comparison of 50
predicted and measured Full-scale pile loading tests
deflections (PMT data 45
available)
(PMT data available)
40
Y0pred./Y0meas.=1,11
Y0/B predicted %
35
30 (R2=94%)
25
20
Lock & Dam, Pile 3-12
15 Lock & Dam, Pile 3-13
Roosevelt bridge, Pile 16
10
Longjumeau, Pile TG
5 Longjumeau, Pile TD
0
0 5 10 15 20 25 30 35 40 45 50
Y0/B measured %
of the PMT data of sand in centrifuge and then to methods of construction of P–Y curves. Due to
an approximate prediction of the piles behaviour. the non-availability of the PMT expansion curves,
As shown in Fig. 14, good prediction is to be the comparison was limited to the method of
noticed for small deflections up to 10% of B. Fascicule-62. For all the piles where the PMT
The ratio Ypred.
0 /Ymeas.
0 of the 27 points of data was available, P–Y curves illustrated by
comparison varied within the interval 0.56–2.40 curve 2 in Fig. 1 were defined and input in
with a mean value of 1.43 and a coefficient of SPULL. Figure 15 summarises the comparison
variation of 30%. between predicted and measured deflections. It
can be seen that the ratio Ypred.
0 /Ymeas.
0 of the 53
points of comparison fluctuates around 0.81
7 Comparative study within a margin of 0.31–3.30 and a coefficient of
variation of 40%. The proposed method slightly
The predictive capability of the proposed method overpredicted the pile deflections whereas the
is to be compared with that of the current method of Fascicule-62 underpredicted them.
70 Y0pred./Y0meas.=1,33
60
(R2=94%)
50
40
30
123
Geotech Geol Eng (2007) 25:283–301 297
Fig. 14 Comparison of 50
predicted and measured
45
deflections (centrifuge
tests) 40
Y0/B predicted %
35
30
25
Centrifuge tests
20
Fontainebleau sand, 1999
15 Le-Rheu sand, 1987 (test 1) Pile ENSM-1
Le-Rheu sand, 1987 (test 2) Pile ENSM-2
10
Le-Rheu sand, 1988 (test 2) Pile P1-2
5 Le-Rheu sand, 1988 (test 4) Pile P1-4
Le-Rheu sand, 1991 (test t) Pile P1-t
0
0 5 10 15 20 25 30 35 40 45 50
Y0/B measured %
70 Y0pred,/Y0meas,=0,81
60 Lock &Dam Pile3-12
R2=84% Lock &Dam Pile3-13
50 Roosevelt bridge Pile 16
Longjumeau Pile TG
40
Longjumeau Pile TD
30 Chatenay Pile T5
Chatenay Pile T10
20 Chatenay Pile T15
Le-Rheu Pile P1
10
Le Rheu Pile P2
0
0 10 20 30 40 50 60 70 80 90 100
Y0 measured (mm)
123
298 Geotech Geol Eng (2007) 25:283–301
123
Geotech Geol Eng (2007) 25:283–301 299
123
300 Geotech Geol Eng (2007) 25:283–301
Bouafia A (2002a) Analysis of lateral reaction modulus for Garassino A (1976) Soil modulus for laterally loaded piles.
piles in sand from CPT test. In: Proceedings of the In: Proceedings of the 4th European Conference on
international symposium PARAM’02, Paris, 2–3 Sep- Soil Mechanics and Foundations Engineering, Vienna,
tember 2002, Presses de l’ENPC 1976
Bouafia A (2002b) Response of a flexible pile under lateral Georgiadis M, Anagnastopoulos C, Saflekou S (1992)
loads in dense sand in centrifuge. In: Canadian Centrifugal testing of laterally loaded piles. Can
Geotechnical Society (ed) Proceedings of ICPMG’02 Geotech J 259:208–216
international conference on physical modelling in Jézéquel JF (1988) Résistance latérale des pieux Le-Rheu-
geotechnics, St-John’s, Newfoundland, Canada July Prévision du comportement des pieux (in French).
10–12 2002c Prelimnary testing report, code F.A.E.R 1.05.01.7,
Bowles JE (1997) Foundation analysis and design, 5th edn. LRPC de St-Brieuc, January 1988, 43 p
Mc Graw-Hill, New York King GJW (1994) The interpretation of data from tests on
Briaud JL (1986) Pressuremeter and foundation design. In: laterally loaded piles. In: Proceedings of the interna-
Clemence SP (ed) Proceedings of IN-SITU’86 use of tional conference CENTRIFUGE’94 on physical
in-situ tests in geotechnical engineering, Geotechnical modelling in centrifuge, 31 August–2 September,
Special Publication No. 6, Virginia Tech., Blacksburg, Singapore, pp 515–520
pp 74–115 MELT (Ministère de l’équipement, du logement et du
Briaud JL, Smith TD, Meyer B (1982) Calcul des pieux transport) (1993) Règles techniques de conception te
chargés latéralement à l’aide des résultats press- de calcul des fondations des ouvrages de génie civil,
iométriques (in French). In: Proceedings of the Fascicule-62, titre-5 (in French), Paris, 182 p
international symposium PMT and its marine appli- Ménard L, Gambin M, Bourdon G (1969) Méthode
cations, Paris, 19–20 April 1982, editions LCPC-IFP, générale de calcul d’un rideau ou pieu sollicité
pp 389–406 horizontalement en fonction des résultats press-
Briaud JL, Smith TD, Tucker LM (1985) A pressuremeter iométriques (in French), Sols/Soils vol 1, No. 20/23,
method for laterally loaded piles. In: Proceedings of 1969, pp 16–28
the international conference on soil mechanics and Ménard L (1971) Le tassement des fondations et les
foundation engineering, San Francisco 1985, vol 3, pp techniques pressiométriques- Bilan après 10 de résul-
1353–1356 tats expérimentaux (in French), Annales de l’ITBTP,
Briaud JL, Moore BH, Mitchell GB (1989) Analysis of pile décembre 1971
load tests at Lock and Dam 26. In: FH Kulhaway (ed) Meyer BJ, Reese LC (1979) Analysis of single piles under
Proceedings of the congress foundation engineer- lateral loading. Research report 244–1, Project 3-5-78-
ing—current principles and practices, Evanston, June 244, Centre for transportation research, The Univer-
25–29, 1989, Evanston, Illinois, vol 2, ASCE, pp 925–942 sity of Texas, 157 p
Canépa Y, Depresle D, Leipp J, Smirr JL (1987) Essais de Mezazigh S (1995) Etude expérimentale de pieux chargés
sollicitations horizontales de tubes de différents latéralement: Proximité d’un talus et effet de groupe
diamètres fichés dans une fosse de sable de Fontaine- (in French). Doctorate thesis, University of Nantes,
bleau (in French). Testing report to the LREP, Code France, 272 p
F.A.E.R 1.15.06.6, file 8670, 32 p Olham D (1983) Lateral loads tests on piles. MSCE thesis,
Carayannacon-Trézos, Baguelin F, Frank R (1979) Réac- University of Manchester
tion latérale des pieux: effets de forme et effets O’Neill MW, Murchisson JM (1983) An evaluation of P-Y
tridimensionnels (in French)’’. Bulletin des LPC No. relationships in sands, Report to the American
104, Nov/Dec 1979 Petroleum Institute (PRAC 82–41–1), May 1983,
Douglas DJ, Davis EH (1964) The movements of buried research report No. GT-DF02–83, 122 p
footings due to moment and horizontal load and the Poulos HG (1971) Behaviour of laterally loaded piles:
movement of anchor plates. Geotechnique, vol 14 I-Single piles, In: Proceedings of the American
Dunand M (1981) Etude expérimentale du comportement Society of Civil Engineers, vol 97, No. Sm5, May
des fondations soumises au renversement (in French). 1971, pp 711–731
Doctorate thesis, University of Grenoble IMG Poulos HG, Carter JP, Small JC (2001) Foundations and
Duncan JM, Chung-Chang (1970) Nonlinear analysis of retaining structures-Research and practice. Report to
stress and strain in soils. In: Proceedings of the ASCE. the 15th international conference on soil mechanics
J Soil Mech Found Div 96:1629–1654 and geotechnical engineering, 27–31 August 2001,
Frank R (1984) Etudes théoriques des fondations profon- Istanbul, vol 4, pp 2527–2606
des et d’essais en place par autoforage dans les LPC et Reese LC (1971) The analysis of piles under lateral
résultats pratiques (in French). Research report to the loading. In: Proceedings of the symposium interac-
LCPC, No. 128 tion structure-foundation, University of Birmingham,
Frank R, Jézéquel JF (1989) La résistance latérale des 1971
pieux (in French), LCPC’s days on Soil Mechanics, Reese L, Matlock H (1977) Non-dimensional solutions for
Saint-Brieuc, 20–22 June 1989, 36 p laterally loaded piles with soil modulus assumed
Gambin M (1979) Calculation of foundations subjected to proportional to depth. In: Proceedings of the 8th
horizontal forces using pressuremeter data. Sols-Soils Texas conference on Soil Mechanics and Foundations
No. 30/31 1979, pp 17–59 Engineering, Austin, September 1956
123
Geotech Geol Eng (2007) 25:283–301 301
Remaud D (1999) Pieux sous charges latérales-Etude Townsend F, Mc Vay (1997) Prediction and evaluation of a
expérimentale de l’effet du groupe (in Franch). laterally loaded pile group at Roosevelt bridge.
Doctorate thesis, University of Nantes, France, 303 p Report submitted to Florida department of transpor-
Robertson PK, Hughes JMO, Campanella RG (1984) tation, No. WPI 0510663, March 1997, 381 p
Design of laterally loaded displacement piles using a Smith TD (1987) Friction mobilisation F-Y curves for
driven pressuremeter. In: Laterally loaded deep laterally loaded piles from the pressuremeter. In:
foundations—analysis and performance, ASTM Spe- Proceedings of the international symposium on pre-
cial Technical Publications No. 835, 1984, pp 229–238 dictions and performance in geotechnical engineering,
Robertson PK, Campanella RG, Brown PT (1985) Design Calgary, Canada, 17–19 June 1987, A. A. Balkema,
of axially and laterally loaded piles using in-situ tests: Amsterdam, pp 89–95
a case history. Can Geotech J 22:518–527
123