Two-Objective Optimization of Aerodynamic Shapes Using Gravitational Search Algorithm

Gravitational search algorithm (for the first time) has been used for two-objective optimization of airfoil shape, in this article. 2D compressible Navier-Stokes equations with Spalart-Allmaras model has been used to simulate viscous and turbulent flow. First, efficiency and accuracy of the optimizer sets have been evaluated using inverse optimization. Objective functions were differences between drag and lift with their corresponding values of the NACA0012 objective airfoil, as a set of airfoils randomly were chosen as starter airfoils, in this case and the aim was to obtain the airfoils that satisfy the considered objective functions. In direct optimization, gravitational search algorithm that has been used in the present work, has achieved proper parameters (related to the Parsec method) and consequently has found optimized airfoils with maximum lift and minimum drag objective functions. This algorithm starts to slove using a set of airfoils and it is directed towards the airfoils that provide the mentioned objective functions. Comparison of the results (Pareto fronts) shows better and more proper performance of the gravitational search algorithm rather than particle swarm optimization algorithm and former researches (done using other meta-heuristic algorithms) for aerodynamic optimizations.