This body of work encompasses a unified treatment of uncertainty associated with multiaxial fatigue damage accumulation. The first part of this study presents a framework for fatigue monitoring on linear components of a structure subjected to unknown loads, using a limited number of output-only vibration measurements. It exploits virtual sensing techniques to estimate the dynamical responses, including virtual multiaxial strain/stress time histories. Joint input-state estimation algorithms are applied to predict the dynamical response of the structural components. The resulting virtual multiaxial stress histories are used to compute the fatigue load cycles and corresponding fatigue damage. The proposed procedure can be adopted over a monitoring period to produce a global fatigue damage...[ Read more ]

This body of work encompasses a unified treatment of uncertainty associated with multiaxial fatigue damage accumulation. The first part of this study presents a framework for fatigue monitoring on linear components of a structure subjected to unknown loads, using a limited number of output-only vibration measurements. It exploits virtual sensing techniques to estimate the dynamical responses, including virtual multiaxial strain/stress time histories. Joint input-state estimation algorithms are applied to predict the dynamical response of the structural components. The resulting virtual multiaxial stress histories are used to compute the fatigue load cycles and corresponding fatigue damage. The proposed procedure can be adopted over a monitoring period to produce a global fatigue damage accumulation map from limited vibration data. The next part of this study proposes a Bayesian framework to re-formulate a multiaxial fatigue model and produce probabilistic stress-life fatigue curves from experimental data. The framework is employed to identify the experimentally-driven parameters that govern the multiaxial fatigue model, in the form of probability distributions. Classical and hierarchical Bayesian inference strategies are presented, accompanied by rigorous analytical expressions for calculating the joint posterior distributions necessary for in-field implementation. This probabilistic treatment makes the existing fatigue models suitable for exercising uncertainty propagation for reliability analysis and design purposes. Lastly, a framework for multiaxial fatigue reliability assessment is proposed to incorporate uncertainty from diverse sources. It thus provides realistic fatigue damage accumulation information for a component based on the actual operational conditions of a structure.