Effective age models for imperfect maintenance

A large number of maintenance models are available in the literature. Most of these usually assume that the effect of maintenance interventions is as good as new. This amounts to assuming the maintenance action undergone by a system, be it preventive or corrective, is equivalent to its replacement. This hypothesis is, of course, questionable in many cases. Maintenance without replacement can lead to a significant level of rejuvenation of a system, either preventively or after repair. However, the restoration of the performances of the system is most of the time incomplete. The effect of such an imperfect maintenance has been described in different ways, which can be split in two main categories: reduction of the value of a degradation variable embodying the ‘health’ of the system, or modification of the lifetime distribution of the system.

This paper focuses on the latter approach, and is structured in two parts. First, it reviews different approaches of imperfect maintenance, modelling the gain in residual lifetime either by a decrease of the failure rate value or by a reduction of the system effective age. Second, an innovative model based on the concepts of elasticity and inescapability of aging is described, in order to introduce more intuitive observations on the results of repeated maintenance actions on a system. This new approach is illustrated using a numerical example.