An optimal design technology for aerodynamic configurations based on the numerical solutions of the full Navier–Stokes equations

A new approach for optimization of essentially 3D aerodynamic shapes intended to minimize their drag is proposed. The method allows one to apply the non-linear surfaces, which are typically used for the complex aircraft junctions such as wing-to-body fairing. This method involves the solution of full Navier–Stokes equations for the objective function calculation, and the optimization technology is based on the genetic algorithms. The important features of the method are both the ability to take into account the multiple geometrical and aerodynamic constraints and the high level of computational efficiency achieved by the complex multilevel parallelization and reduced-order modeling approach. This method was applied for a wing-tobody fairing optimization on a typical medium-range aircraft at the realistic transonic flight conditions.
The constraint handling can be described as follows: it is proposed to employ the search paths through both feasible and infeasible points instead of a traditional approach where only feasible points can be included in a path. For this purpose, the search space is extended by evaluating points (in terms of fitness) which do not satisfy the constraints imposed by optimization problem. The required extension of an objective function can be implemented due to a basic property of genetic algorithms: they are not confined to only smooth solutions in contrast to the classical optimization methods.
The results demonstrate that the proposed approach provides a significant drag reduction and is applicable for engineering and designing.