Best practices for surrogate based uncertainty quantification in aerodynamics and application to robust shape optimization

Abstract

This chapter introduces the use of aerodynamic shape optimization applied to industrial problems, motivates the use of a robust approach over the classical deterministic optimization, and presents different alternatives for the robust-based and reliability-based problems. The use of surrogates for the Uncertainty Quantification of operational and geometrical uncertainties is a cost-effective solution for high dimensional models if the gradient information is introduced by means of the adjoint method. Finally, the proposed methodology is applied through the reliability-based optimization of an airfoil under operational uncertainties.