Abstract: Shape optimization of the airfoil with high aspect ratio
of long endurance unmanned aerial vehicle (UAV) is performed by the
multi-objective optimization technology coupled with computational
fluid dynamics (CFD). For predicting the aerodynamic characteristics
around the airfoil the high-fidelity Navier-Stokes solver is employed
and SMOGA (Simple Multi-Objective Genetic Algorithm), which is
developed by authors, is used for solving the multi-objective
optimization problem. To obtain the optimal solutions of the design
variable (i.e., sectional airfoil profile, wing taper ratio and sweep) for
high performance of UAVs, both the lift and lift-to-drag ratio are
maximized whereas the pitching moment should be minimized,
simultaneously. It is found that the lift force and lift-to-drag ratio are
linearly dependent and a unique and dominant solution are existed.
However, a trade-off phenomenon is observed between the lift-to-drag
ratio and pitching moment. As the result of optimization, sixty-five
(65) non-dominated Pareto individuals at the cutting edge of design
spaces that is decided by airfoil shapes can be obtained.
Abstract: Predictions of flow and heat transfer characteristics and shape optimization in internally finned circular tubes have been performed on three-dimensional periodically fully developed turbulent flow and thermal fields. For a trapezoidal fin profile, the effects of fin height h, upper fin widths d1, lower fin widths d2, and helix angle of fin ? on transport phenomena are investigated for the condition of fin number of N = 30. The CFD and mathematical optimization technique are coupled in order to optimize the shape of internally finned tube. The optimal solutions of the design variables (i.e., upper and lower fin widths, fin height and helix angle) are numerically obtained by minimizing the pressure loss and maximizing the heat transfer rate, simultaneously, for the limiting conditions of d1 = 0.5~1.5 mm, d2 = 0.5~1.5 mm, h= 0.5~1.5mm, ? = 10~30 degrees. The fully developed flow and thermal fields are predicted using the finite volume method and the optimization is carried out by means of the multi-objective genetic algorithm that is widely used in the constrained nonlinear optimization problem.