Multi-Objective Aerodynamic Optimization of an Unmanned Aerial Vehicle
2016 (English)Conference paper, Presentation (Other academic)
In the present work, a global multi-objective optimization methodology with meta-models is developed and applied for optimizing the aerodynamic performance of an unmanned aerial vehicle (UAV) airfoil. The shape of the airfoil is obtained using a parameterization scheme based on Bezier curves. The aim is to optimize the performance of a high-lift airfoil by increasing the lift force while minimizing the pitching moment generated by the integration of the viscous and pressure forces acting on the airfoil. The proposed strategy for determining the optimal design solution for maximum lift and minimum moment is based on the construction of meta-models in conjunction with robust optimization algorithms. In more detail, a direct multi-objective optimization methodology using the concept of Pareto front is utilized. A relatively small number of design points is evaluated and their function values are stored in a database for meta-model construction. Radial Basis Functions such as multi-quadrics are used for the meta-model construction. The overall optimization process is integrated with Computation Fluid Dynamics techniques, using automatic schemes that build on parameterized airfoil geometries to construct a suitable mesh through scripts. Tcl/Tk language is employed along with the commercial software ICEM-CFD in batch mode. For the numerical calculation of the flow, the software FLUENT is used with fluid properties, turbulence model and boundary conditions set also in batch mode. It is demonstrated that the proposed methodology allows to develop realistic high performance airfoil conceptual designs, with relatively little computational effort. The methodology is generic and can be applied in a variety of problems related to fluid mechanics and heat transfer.
Place, publisher, year, edition, pages
Multi-objective Optimization, Meta-model, Airfoil, Computational Fluid Dynamics, Aerodynamics, Radial Basis Functions, Pareto-front
IdentifiersURN: urn:nbn:se:mdh:diva-33521OAI: oai:DiVA.org:mdh-33521DiVA: diva2:1045397
Aerospace Technology Congress - FT2016