62
Geotechnical News June 2011
THESIS ABSTRACTS
der of seconds rather than hours, as for MLTs. In addition, offshore
structures are usually designed for events that may not occur for a
significant period of time after installation and therefore the impact
of time on pile capacity needs to be considered. This thesis presents
an experimental investigation of these factors with a view toward
incorporating these biases into the design of offshore piles.
To investigate the impact of end condition, both open-ended
and closed-ended pile tests were performed. The design, develop-
ment and construction of these piles were carried out such that they
were representative of offshore geometries but at a reduced scale
for maneuverability and testing requirements. The end condition
was shown to have a significant impact on the base resistance and a
lesser impact on the shaft resistance. The base resistance was seen
to increase directly with the degree of plugging experienced during
installation. In contrast the unit shaft resistance was relatively inde-
pendent of the base geometry with the closed and open-ended piles
mobilizing similar radial effective stresses during installation. The
long term shaft resistance suggested minimal impact of end condi-
tion, with the open-ended pile mobilizing comparative stresses to
the closed-ended pile. These results suggest that design equations
formulated from closed-ended tests can be used to predict open-
ended shaft capacities in normally consolidated clays.
The cyclic load tests demonstrated a marked transition from
stable to unstable behaviour for loads in excess of 74% of the stat-
ic capacity. Unstable behaviour was characterized by rising pore
pressures and decreasing effective stresses alongside accelerating
displacements. Rapidly applied CRP loads yielded capacities well
above the estimated static capacities. Interface dilation was shown
to control the resistance to rapid loading, with the capacity increas-
ing dramatically in response to decreasing shear induced pore pres-
sures, which typically occurred after 2.5-4.0 mm pile head move-
ment. Similar depressed pore pressures were observed over the
initial 10 cycles of load during unstable cycling. The implication for
design is that offshore piles can withstand single rapid events in ex-
cess of the static capacity, or can withstand a limited number of high
level cycles above the loading threshold but cannot sustain continu-
ous high level cycling above approximately 74% of the capacity.
The reload testing conducted on the 250mm square concrete
piles resulted in capacities which were 50% higher than the initial
virgin load tests, suggesting a positive ageing effect. Comparison
between these tests and the literature showed good agreement. The
potential for incorporating time effects into the design process is
demonstrated through a simple reliability analysis which highlights
the improved reliability of piles with time.
Supervisor: Dr. Kenneth Gavin, University College Dublin
Automatic Generation of Solid Models of
Building Façades from Lidar Data for Com-
putational Modelling
Linh Truong-Hong
Linh Truong-Hong, Room G67, Newstead Building, Urban Model-
ling Group, School of Architecture, Landscape and Civil Engineer-
ing, University College Dublin, Belfield, Dublin 4, Ireland, Phone:
+353 1716 3232, email:
For assessment of building damage, geometric models of exist-
ing buildings are required for analyzing processing by using ad-
vanced finite element codes. Traditionally, the geometric models
were often created by using manual surveys that are expensive and
time consuming, especially when numerous or complex buildings
are involved. Light Detection and Ranging (LiDAR) technology
provides a new technique to measure surfaces of objects with fast
and high accuracy, but has to date not been effectively exploited as
the basis for computational models.
In this study, the terrestrial laser scanning data were used as the
input for several novel approaches to automatically reconstruct geo-
metric models for computational modeling by employing a voxel-
ized octree representation. Possible boundary lines of openings and
the façade were determined from the boundary points underlying
voxels on their boundaries. A new method called the “flying voxel”
approach was devised to determine voxel position with respect to a
façade and its major features. Finally, all full voxels were converted
into a neutral file describing the solid model for importing directly
finite element codes.
Reliability of the algorithms was verified against measured draw-
ings based on the accuracy of the geometric models and the efficacy
of their response during computational analysis. In term of the phys-
ical geometry, overall dimensions and opening areas of the building
facades were respectively less than 1.5% and 3.0% of relative er-
rors compared to real facades. An average absolute error of opening
dimensions was mostly 33.4mm. In terms of building response, a
macro model with non-linear analysis was adopted to analyze the
buildings subjected to excavation-induced settlements. Responses
deviated by less than 5% of the relative errors in angular distortion
and nodal displacements compared to finite element results obtained
from the real façades. The maximum estimated absolute error of
nodal displacements at nodes of interest was around 4.3mm. Also,
distribution of stress-strain and cracking patterns were graphically
consistent with ones derived form the measured drawings.
Sponsoring Professor: Dr. Debra F. Laefer, Urban Modelling
Group, School of Architecture, Landscape, and Civil Engineering
University College Dublin
Instabilities in Sands
Alfonso Mariano Ramos Cañón
Alfonso Mariano Ramos Cañón, Geophysical Institute Javeriana
University. Cra 7 No 42-27. Bogotá – Colombia,
email:
The purpose of this dissertation is to contribute to the detec-
tion of the onset of geomechanical instabilities in sands under dry/
drained and saturated/undrained conditions. In order to accomplish
this objective, a framework for detecting instabilities based on either
a mathematical concept (i.e., loss of uniqueness - bifurcation) or
physical one (i.e., Hill´s instability) is generated to derived criteria
applicable to the most common instabilities: localized drained in-
stability in dense sands (shear bands), diffuse undrained instability
in loose sands (liquefaction) and diffuse drained instability in loose
and dense sands (debris flow). Each one of these three instabilities
is studied independently. The criteria are compared against experi-
mental results available in the literature, and reasonable agreement
is achieved. From a practical perspective, the contributions of this
work expand the repertoire of potential instabilities that have been
reported in case studies of puzzling slope instability failures under
drained and undrained conditions.
Sponsor: Professor Arcesio Lizcano, University de Los Andes. De-
partment of Civil and Environmental Engineering. Cra 1E No 19A
– 40. Bogotá - Colombia.
