Aerodynamic shape optimization of two-dimensional airfoils under uncertain operating conditions
Huyse Luc, Lewis Robert Michael · 2001
Practical experience with airfoil optimization techniques has revealed unexpected difficulties. Traditionally the performance of an airfoil is optimized for given, or assumed, model parameters. Experience has indicated that a deterministic optimization for discrete operating conditions may result in dramatically inferior performance when the actual conditions are different from these, somewhat arbitrarily chosen, design values. Extensions to multi-point optimization have proven unable to adequately remedy the problem of ``localized optimization''''. This paper presents an intrinsically statistical approach and demonstrates how the shortcomings of multi-point optimization with respect to ``localized optimization'''' can be overcome.