Inspection Planning for Deteriorating Structures and the Influence of Inevitable Uncertain Parameters

Abstract

In the previous paper, the authors presented a sequential cost minimization method for the inspection planning problem of fatigue deteriorating structures. The method aims to find an optimal strategy so that the total expected cost in the period between the present inspection and the next be minimum. The optimization is repeatedly carried out at every inspection. The sequential cost minimization is a sub-optimization method of inspection strategy. However, the strategies predicted by the method are close to the lifetime optimal one in most cases. Further, the optimization can be accomplished with realistic computational effort even for large system.In this paper, first of all, the content of the sequential cost minimization method is briefly reviewed, and the applicability of the method is examined for an actual member set consisting of structural elements with round fillet weld. Then, the influences of inevitable uncertain parameters on inspection planning are discussed.The uncertainties related to inspection planning can be classified into two categories : engineering uncertainty and economical/social uncertainty. With respect to the engineering uncertainty, it is further divided into two groups : reducible uncertainty and non-reducible uncertainty.For reducible uncertainty, initial uncertainty can be gradually reduced by updating the inspection results and/or service experiences. The Bayesian analysis is applied in conjunction with the sequential cost minimization method for a member set which has an uncertainty in the fatigue property of members. It is made clear that the “visual and conditional mechanical inspection method” and “sampling mechanical inspection method” become profitable when a large uncertainty exists in the fatigue life of members.For non-reducible uncertainty, initial uncertainty remains through the whole service of the structures. Two types of sensitivity analysis are carried out. First, giving a small change for individual parameters, the degrees of influence of each parameter on the inspection content, the cumulative operating cost and the failure probability are investigated. Second, sensitivity analysis is carried out by giving a large change for a single parameter. As a result, it is found that the optimization of inspection strategy itself has an effect to decrease the influence of this large uncertainty on the operating cost.