Offshore structures are subject to a wide variety of loads such as
wind, wave and current) all of which exhibit a high degree of
statistical uncertainty. The structural components, e.g. beams,
columns, connections, piles, conductors and risers must be
proportioned to resist the effects of these loads. Uncertainties in
capacities or member strengths arise because of material and
fabrication variability, and limitations in engineering theories to
predict and conveniently interpret element and system response and
capacity. As a result, a reliability-based approach is necessary to
account for these uncertainties.This project aimed to extend the
applicability of recently developed methods : a Factor Analysis
Technique; a Dynamic Excess Kurtosis technique; and Najafian's
Extreme Value probability model (NEV). The specific objectives were
:- to incorporate the effect of current and load intermittency in the
splash zone into the NEV probability modelling approach; and also to
extend the techniques for efficient evaluation of kurtosis (a crucial
parameter in the characterisation of a probability distribution) to
cases when current , load intermittency and wave directionality are
important. Describes a number of achievements: a robust
implementation of the simulation technique, directed towards the
reliable estimation of extreme values of response has been
successfully
extended to account for current and load intermittency in the splash
zone; the Factor Analysis technique has been extended to account for
current, wave directionality and load intermittency in the splash
zone. The results of this technique have been compared against those
from time simulation technique to confirm both its accuracy and its
much higher computational efficiency (about 25 times for long-crested
seas and at least 100 times for directional seas). The effects of
current and load intermittency on the statistical properties of both
dynamic and quasi-static responses have been thoroughly investigated.
Extension of the Dynamic Excess Kurtosis Technique and Najafian's
Extreme Value probability model to the case of structures under the
influence of load intermittency and current has been thoroughly
investigated. Extensive investigation has shown that the
Pierson-Holmes distribution can accurately account for the effect of
current on the probability distribution of response. However, it
cannot accurately model the effect of the load intermittency in the
splash zone on the response probability distribution. A methodology
for establishing an improved probability model for the response
distribution arising from structures subject to load intermittency
has been devised, the proposed methodology can also be used to
determine the probabilistic properties of the dynamic response from
those of the quasi-static response.
Some of these research reports are very large, for information about expected download times please see our FAQs pages.
