Adhikari, S. and Langley, R. S.
Third International Conference on Mathematical Methods in
Reliability (MMR 2002), pp. 3-6, Edited by H. Langseth and B.
Lindqvist, June 17-20, 2002 Trondheim, Norway.
In the reliability analysis of a complex engineering structure a very large number of the system parameters can be considered to be random variables. The difficulty in computing the failure probability increases rapidly with the number of variables, and the aim of the present paper is to consider methods whereby the number of variables can be reduced without compromising the accuracy of the reliability calculation. The most common methods of reliability prediction are FORM (First Order Reliability Method) and SORM (Second Order Reliability Method), and in each case the efficiency of the method depends on the efficient computation of the so called reliability index, commonly denoted by `b'. The geometric interpretation of the reliability index is that it is the minimum distance of the failure surface from the origin in a transformed space (referred as the z-space) where all the random variables defining the uncertainties of the system are Gaussian, normalized and uncorrelated. The present reduction methods are based on the sensitivity of the failure surface in the z-space. If the failure surface is close to linear, then the values of b obtained from these methods are close to the exact value of b obtained using the full set of random variables. However, if the failure surface is significantly nonlinear, the different reduction methods introduce different kind of errors. The nature of these errors are studied using a wide range of numerical examples. It is shown that the values of b obtained using the proposed reduction methods have acceptable accuracy for many large scale structural engineering problems.
@INPROCEEDINGS{cp7,
AUTHOR={S. Adhikari and R. S. Langley},
TITLE={Reduction of random variables in structural reliability analysis},
BOOKTITLE={Proceedings of the third International Conference on
Mathematical Methods in Reliability Methodology and Practice (MMR 2002)},
YEAR={2002},
Editor={H. Langseth and B. Lindqvist},
Pages={3-6},
Address={Trondheim, Norway},
Month={June},
Note={}
}