Discrete Adjoint-Based Simultaneous Analysis and Design Approach for Conceptual Aerodynamic Optimization

摘要

In this paper, a simultaneous analysis and design method is derived and applied for a non-linear constrained aerodynamic optimization problem. The method is based on the approach of defining a Lagrange functional based on the objective function and the aerodynamic model’s equations, using two sets of multipliers. A fully-coupled, non-linear system of equations is derived by requiring that the Gateaux variation of the Lagrange functional vanishes for arbitrary variations of the aerodynamic model’s dependent variables and design parameters. The optimization problem is approached using a one-shot technique, by solving the non-linear system in which all sensitivities and problem constraints are included. The computational efficiency of the method is compared against a gradient-based optimization algorithm using adjoint-provided gradient. A conceptual-stage aerodynamic optimization problem is solved, based on a non-linear numerical lifting-line method with viscous corrections.