Calibration methods for reliability-based design codes

The calibration methods are applied to define the optimal code format according to some target safety levels. The calibration procedure can be seen as a specific optimization process where the control variables are the partial factors of the code. Different methods are available in the literature, but the choice of an appropriate method is not usually an easy task. Each method can be shown to have a better performance for a particular kind of problems. The scope of this paper is to underline the advantages and disadvantages of the classical methods, to define their domain of validity and to propose new efficient methods. In our approach, the accuracy is balanced with time efficiency by the mean of iterative scheme using approximate methods. The presented ideas are clarified by four numerical examples of parabolic performance function, column buckling, stability of submarine shells and structural elements. (C) 2003 Elsevier Ltd. All rights reserved.