| dc.description.abstract | The degradation of metallic systems under cyclic loading is prone to significant uncertainty. This uncertainty in turn affects the reliability in the prediction of residual lifetime and the subsequent decision regarding the optimum inspection and maintenance schedules. In particular, the experimental data on the evolution of fatigue-induced cracks shows significant scatter stemming from initial flaws, metallurgical heterogeneities, and randomness in material properties like yield stress and fracture toughness. The objective of this research is to improve the reliability-based optimal inspection planning of metallic systems subjected to fatigue, taking into account the associated uncertainty. To that end, this research aims to address the two main challenges faced in developing a credible reliability-based framework for lifecycle management of fatigue-critical components. The first challenge is to construct a stochastic model that can adequately capture the nonlinearity and uncertainty observed in the crack growth histories. The second one involves presenting a computationally efficient strategy for solving the stochastic optimization associated with optimum maintenance scheduling. In order to fulfill these objectives, a Polynomial Chaos (PC) representation is constructed of fatigue-induced crack growth process using a database from a constant amplitude loading experiment. The PC representation relies on expanding the crack growth stochastic process on a set of random basis functions whose coefficients are estimated from the experimental database. The probabilistic model obtained is then integrated into a reliability framework that minimizes the total expected life-cycle cost of the system subjected to constraints in terms of time to inspections, and the maximum probability of failure defined by the limit state function. Lastly, an efficient and accurate optimization strategy that uses surrogate models is suggested to solve the stochastic optimization problem. The sensitivity of the optimum solution to the level of risk is also examined. This research aims to provide a decision support tool for informed decision-making under uncertainty in the life-cycle planning of systems subjected to fatigue failure. | en |
