This dissertation presents an integrated framework that systematically captures and propagates uncertainties in advanced composite materials and aerodynamic loads to assess the structural reliability of unmanned aircraft. The research introduces a nov...
This dissertation presents an integrated framework that systematically captures and propagates uncertainties in advanced composite materials and aerodynamic loads to assess the structural reliability of unmanned aircraft. The research introduces a novel combination of R-vine copula modeling, Sparse Polynomial Chaos Expansion (SPCE), and Bayesian inference to quantify composite material variability from the microstructural level to full-scale laminate properties.
For loads with inherent uncertainties, the proposed finite element analysis (FEA) strategy models element-level aerodynamic loads using Gaussian Random Fields (GRFs) that reflect probability distributions and spatial correlations. This approach enables the integration of deterministic safety factors with stochastic load modeling, ultimately deriving failure probabilities and failure rates for critical maneuvers, which form the basis for structural reliability assessments aligned with airworthiness standards.
Furthermore, the framework includes a dynamic impact analysis for bird strike scenarios using Smoothed Particle Hydrodynamics (SPH), enabling a precise representation of localized, high-intensity pressure effects on overall structural integrity. In addition, a design strategy is proposed that selectively reduces safety factors for primary and secondary structures under combined dynamic impact and maneuvering conditions, thereby achieving structural weight reduction while still satisfying airworthiness requirements.
The proposed framework was validated using a full-scale unmanned aircraft model, where it defined and quantified the uncertainties and variabilities in both composite materials and loads, thus demonstrating high performance, computational efficiency, and strong potential for industrial application. Ultimately, this dissertation presents an integrated methodology for unmanned aircraft assessment that is structurally reliable, efficient, and compatible with airworthiness certification frameworks.