Entropy-based framework for optimal life-cycle management of multiple fatigue cracks in ship structures

Abstract

Cyclic loading of ship structures can result in multiple cracks at fatigue-prone details. While these cracks are generally considered a nuisance, they can result in catastrophic failure if left unrepaired. On the other hand, periodic inspection and repair of these cracks can result in a large financial burden. Accordingly, careful scheduling of inspection and repair of the cracks is needed to avoid catastrophic failures while ensuring low cost. Life-cycle analysis offers a reasonable means to establish cost-effective scheduling of maintenance and repair strategies for civil and marine infrastructure. Studies exist on life-cycle cost frameworks for the assessment and repair of multiple fatigue cracks utilizing non-linear optimization strategies. In this study, we establish a new probabilistic life-cycle cost framework that improves on existing approaches by utilizing the concept of entropy to determine optimal inspection and repair scheduling for ship structures with multiple cracks. The parameters influencing the propagation rate of the cracks are considered random variables with predefined statistical distributions. A parametric study is conducted to assess the impact of some parameters, such as design life and loading cycles on the obtained life-cycle cost. The results show stark improvement in achieving optimal scheduling of inspection and repair with low cost over traditional optimizations.