Metamodeling fluid systems is an efficient way to do complex analysis on multivariate problems and can allow for time savings in an optimization setting. Gaussian process metamodels provide a flexibility in modeling that can be extended to both experimental and deterministic experiments. This paper specifically addresses such models applied to computational fluid dynamics analysis of a D circulation control wing. The framework for a generalized approach is first overviewed and then applied to the analysis of the aerodnamics. The Gaussian process regression was essential for both simplifying the calculations required for balanced field length computation and for analysis of complex multi-variable behavior.


Aerospace Engineering

Publisher statement

The definitive version is available at http://www.aiaa.org/content.cfm?pageid=230&lumeetingid=1820#zz1820.



URL: http://digitalcommons.calpoly.edu/aero_fac/65