Structure dynamic reliability: A hybrid approach and robust meta-models

Mathematical modeling of physical systems is essential to understand and, if possible, control such systems. However, insufficient information may be available about the level of uncertainty related to material properties, geometric parameters, boundary conditions and the applied loads. In the context of structural reliability, the uncertainties may be uncontrollable when designing for robustness. These problems in the modeling of the uncertainties are often complicated by the model's inability to describe the physical phenomena that are involved. In this paper, the proposed approach combines a dynamic reliability method and a meta-model (reduced model) to obtain good results in terms of the reliability and optimization of such systems. Using the available information about the uncertain design parameters, we use the hybrid model coupling of the possibility and probability approaches for the propagation of the uncertainties in the model. The proposed method was implemented on theoretical structures with different meta-models. The results are compared with the Monte Carlo simulations. This allowed us to prove the robustness and efficiency of the proposed methodology for reliability calculations of complex dynamic structures.