Page 437 - Marine Structural Design

Part IV

Structural Reliability

Chapter 23 Basics of Structural Reliability

23.1 Introduction
Part IV describes structural reliability methods for the design of marine structures with
emphasis on their practical application, e.g. to ship structures. Focuses are given to basic
concept, methodology and application. Examples are given to demonstrate the application of
the methodology.
Details of the structural reliability theory can be referred to, e.g. Ang and Tang (1975, 1984),
Thoft-Christensen and Baker (1982), Madsen (1986), Schnerder (1997), Melchers (1999).
Discussions are given on simple analytical equations that are based on lognormal assumptions.
The papers on numerical approaches, e.g. Song and Moan (1 998) are also mentioned briefly.
The following subjects are addressed in detail:
Reliability of marine structures
Reliability based design and code calibration
Fatigue reliability
Probability and risk based inspection planning

23.2 Uncertainty and Uncertainty Modeling
23.2.1 General

In general, a marine structural analysis deals with the load effects (demand) and the structural
strength (capacity). In design, the dimensions of the structural members are determined based
on the requirement that there is a sufficient safety margin between the demand and the
capacity.
Uncertainties are always involved in all the steps of structural analysis and in strength
evaluation. These uncertainties are due to the random character of the environment, geometric
and material properties, as well as inaccuracy in prediction of loads, response and strength.
Rational design and analysis of marine structures require consideration of all the uncertainties
involved in predicting load effects and structural modeling. Uncertainty analysis is the key in
any reliability evaluation such as reliability-based design and re-qualification for marine
structures.
The development of probabilistic analysis methods and design codes increased the importance
of quantifylng uncertainties. The results of the studies on uncertainty modeling can be used to