Page 439 - Marine Structural Design
P. 439

Chapter 23 Basics of Structural Reliability                           415

                 Equation (23.2) is valid for small coefficients of variation (less than 0.10) only. However, the
                 above approximations are frequently used.

                 23.3  Basic Concepts

                 23.3.1  General
                 Structural engineering deals with load (S) and strength (R) in terms of forces, displacements
                 and stresses acting on the structures. Structural design codes commonly specify loads, strength
                 and appropriate safety factors to be used. Structural reliability theory is about the evaluation of
                 the failure probability taking into account the uncertainties in loads and strength. During the
                 last two decades, many efforts have been given on structural reliability and their application to
                 practical structural engineering.
                 23.3.2  Limit State and Failure Mode
                 A structural component can fall into safe or failure state. The border line (or surface) between
                 the safe and failure states is named as limit state, and expressed as g(Z) = R -S  . The following
                 conditions describe the possible states of a structural component.
                       g(Z)<O represents a failure state where loads S exceeds the strength R.
                       g(Z)>O represents a safe state since strength R is larger than loads S.
                       g(Z)=O represents the limit state line (or surface).
                 The figure below shows the concept of limit state sketchily
                 For  marine  structures,  the  limit  states  are  defined  in  accordance  with  the  different
                 requirements, such as serviceability, ultimate strength, etc.
                 23.3.3  Calculation of Structural Reliability
                 By  quantifying the  uncertainties using  probabilistic methods,  Structural  reliability can  be
                 measured by means of failure probability.
                 For a structure described by a set of random variables Z with joint distribution fi(z), it must be
                 possible for each set of values of z to state whether or not the structure has failed. This leads to
                 a unique division of Z space into two sets, calIed the safe set and the failure set respectively.
                 These two sets are separated by the failure surface (limit state).
   434   435   436   437   438   439   440   441   442   443   444