Robust hybrid approach to multiobjective constrained optimization in aerodynamics

A new approach to the multiobjective constrained design of aerodynamic shapes is suggested. The approach employs genetic algorithms (GA) as an optimization tool in combination with a reduced-order-models method based on linked local databases obtained by full Navier-Stokes computations. The method was applied to the problem of multiobjective transonic profile optimization with nonlinear constraints. The results demonstrated that the method retains high robustness of conventional GAs while keeping computational-fluid-dynamics computational volume to an acceptable level as a result of a limited use of full Navier-Stokes computations. A significant computational time saving (in comparison with optimization tools fully based on Navier-Stokes computations) allowed the algorithm to be used in a demanding engineering environment.