Bayesian Sampling using Condition Indicators

Authors

Michael H. Faber and John D. Sørensen

Abstract

The problem of control quality of components is considered for the special case where the acceptable failure rate is low, the test costs are high and where it may be difficult or im-possible to test the condition of interest directly. Based on the classical control theory and the con-cept of condition indicators introduced by Benjamin and Cornell (1970) a Bayesian approach to quality control is formulated. The formulation is then extended to the case where the quality con-trol is based on sampling of indirect information about the condition of the components, i.e. condi-tion indicators. This allows for a Bayesian formulation of the indicators whereby the experience and expertise of the inspection personnel may be fully utilized and consistently updated as fre-quentistic information is collected. The approach is illustrated on an example considering a con-crete structure subject to corrosion. It is shown how half-cell potential measurements may be util-ized to update the probability of excessive repair after 50 years.

Published in/by

Proceedings ICOSSAR '01, 8th International Conference on Structural Safety and Reliability, eds. R.B. Corotis, G.I. Schuëller and M. Shinozuka, Newport Beach, CA, USA, June 17-22, 2001.