Multifidelity Robust Aeroelastic Design

Nielsen Engineering & Research (NEAR) proposes a new method to generate mathematical models of wind-tunnel models and flight vehicles for robust aeroelastic analysis and design. These models provide a unified description applicable to CFD steady and unsteady aerodynamics, reduced-order CFD approaches, flexible structures and active control systems, and can accommodate probabilistic aerodynamics and aeroelastics. NEAR's offering is based on probabilistic metamodels which are supported by analyses and data at all available levels of fidelity and which are dynamically updated based on multifidelity expected improvement concepts. The proposed software will help reduce the design and life-cycle cost of next-generation high-efficiency flight vehicle systems and revolutionary aerospace vehicles, and will help attain better aeroelastic designs, by providing a better understanding of how the design variables interact and affect each other under the influence of uncertainty, and by incorporating these interactions early in the design to reduce risk.