Time-dependent analyses present special challenges for multidisciplinary design and analysis, and the ability to include multifidelity representations of these phenomena, including CFD unsteady aerodynamics, must be considered in future analyses.
The objective of this work was to develop and evaluate the feasibility of using efficient recurrence frameworks to model unsteady aeroelastic responses in a way that is compatible with recently developed multifidelity optimization methods.
Various algorithms and methods were developed for forming NARMAX models of aeroelastic systems and for measuring their performance.
This framework was successfully applied to the simulation of an AGARD 445.6 wing wind-tunnel model undergoing forced small amplitude pitch oscillations at transonic speeds.
Compact recurrent model representations were achieved with good Model Predicted Output accuracy.
The effects of sampling, nonlinearity, sensitivity to noise, and phase error were investigated, and the scalability of the method, including the incorporation of variable fidelity data, was analyzed.
The results suggest the viability of this approach for general and potentially large-scale aeroelastic applications, providing the foundation for a cost-effective framework for multifidelity robust aeroelastic design.
Reisenthel, P. H., Lesieutre, D. J., and Quijano, O. E., "Robust Recurrence Framework for Unsteady Multifidelity Aeroelastic Analysis," NEAR TR 670, Nielsen Engineering & Research, Santa Clara, CA, Sep. 2011.