2803
Bored pile foundation response using seismic cone test data
Réponse des pieux à l'aide des données de piézocône sismique
Mayne P.W.
Georgia Institute of Technology, Atlanta GA USA
Woeller D.J.
ConeTec Investigations, Richmond, BC Canada
ABSTRACT: A closed-form elastic continuum solution is used to represent the upper and lower segment response of bored piles
subjected to bi-directional Osterberg load testing. For geotechnical parameter input, seismic piezocone tests (SCPTu) are a most
efficient and economical means because the penetrometer readings provide data for assessing the capacity of side and base
components, while the shear wave velocity provides the fundamental stiffness for displacement analyses. A simple algorithm for
modulus reduction is employed to allow nonlinear load-displacement-capacity behavior. A load test case study involving two levels of
embedded O-cells for a large bridge in Charleston, South Carolina is presented to illustrate the approach.
RÉSUMÉ: Une solution analytique en élasticité est utilisée pour représenter la réponse des sections supérieures et inférieures de
pieux forés soumis à des essais de chargement bidirectionnel Osterberg. Les essais sismiques utilisant des piézocônes (SCPTu) sont
des moyens efficaces et économiques pour obtenir des paramètres géotechniques. Les mesures du pénétromètre sont généralement
utilisées pour l’analyse du frottement latéral et de la capacité portante, tandis que les données de vitesse des ondes de cisaillement
donnent le module de cisaillement initial lors de l’analyse des déplacements. Un algorithme simple pour la réduction du module est
utilisé pour l’analyse non-linéaire du comportement charge - déplacement - capacité. Une étude de cas sur un essai de chargement
Osterberg à deux niveaux effectué sur un grand pont situé à Charleston en Caroline du Nord est utilisée afin d’illustrer cette
démarche.
KEYWORDS: bored piles, cone penetration, deep foundations, elastic solution, load tests, shear modulus, shear wave velocity
1 INTRODUCTION
The Osterberg load cell provides a novel alternative to
conventional static load tests that either rely on large dead
weight frames or anchored reaction beams. The O-cell occupies
a minimal space in its setup, essentially the same space taken by
the bored pile foundation itself (Osterberg 1998, 2000). In the
original design, a single sacrificial hydraulic jack is situated at
the pile base (O'Neill, et al. 1997). After concrete placement and
curing, the jack is pressurized resulting in bi-directional loading
to simultaneously mobilize end bearing downward while
pushing the shaft segment upward (Fellenius, 2001). After
loading, the O-cell is grouted to become part of the completed
and working foundation. In fact, the O-cell can be positioned at
mid-level elevations within the pile shaft to better match and
optimize opposing segments. Moreover, multiple O-cells are
now utilized to stage load separate pile segments and verify
higher load capacities.
An elastic pile solution (Randolph and Wroth, 1978, 1979)
is shown to accommodate the various O-cell configurations, as
well as the more common top-down loading of bored piles. The
results from seismic piezocone tests (SCPTu) are shown to be
applicable for providing all the necessary input parameters to
drive the computations and generate curves on the axial pile
deformation response. A case study from South Carolina is used
to illustrate the methodology.
2 ARTHUR RAVENEL BRIDGE
This newly-completed I-17 bridge over the Cooper River in
Charleston, SC was supported by over 400 large bored pile
foundations having diameters of 1.8 to 3 m and embedded
lengths of between 45 to 72 m. The 4.0-km long cable-stayed
concrete segmental bridge has a main span length of 471 m and
connects the city of downtown Charleston with Drum Island
and Mount Pleasant, replacing two old steel truss crossways
known as the Pearman and Grace Memorial bridges.
In the region, all significant heavy building, port, and civil
structures are founded on a deep overconsolidated formation
termed the Cooper Marl that generally lies below elevations -15
to -20 m MSL (Camp, 2004). The uppermost soils consist of
soft variable clays, loose sands, silts, and peats of Holocene age
which are heterogeneous deposits from marine, deltaic, and
alluvial origins
2.1
Cooper Marl
The Cooper marl is a marine deposit consisting of stiff green-
gray sandy calcareous clay of Oligocene age that has been
preconsolidated by erosional processes and natural cementation.
The marl has a high calcite content on the order of 60 to 80 %.
Mean values of indices from laboratory tests include: w
n
=
48%, w
L
= 78%, and PI = 38% (Camp et al. 2002). Typical
SPT-N values in the Cooper marl are between 12 and 16
blows/0.3 m. Triaxial tests consistently show high effective
stress friction angles averaging
'
≈
44º ± 3º. An equivalent
prestress
v
' = (
p
' -
vo
') = 480 kPa captures the general trend
of preconsolidation stress (
p
') which increases with depth, as
well as corresponding profile of overconsolidation ratio (OCR =
p
'/
vo
') that decreases with depth (Mayne 2007a).
For the bridge project, approximately 45 rotary drilled
borings and 55 SCPTu soundings were completed to depths of
55 m. Two representative SCPTu soundings are presented in
Figure 1. The SCPTu is a particularly efficient and economical
means for site exploration as it provides five separate readings
on soil response with depth, including: cone tip resistance (q
t
),
sleeve friction (f
s
), porewater pressures (u
2
), time rate of
dissipation (t
50
), and downhole-type shear wave velocity (V
s
).
1
