Abstract
A Fully Stressed Element approach to topology optimization of an Unmanned Aerial Vehicle wing spar is presented in this paper. The proposed method, a modification of the Fully Stressed Design approach, was applied in optimizing the mass and maximum deflection of a solid, cantilevered wing spar in pure bending, ultimate load condition. This was achieved in two steps. The first step minimized the mass by applying a low resolution, discrete element grid on the structure while imposing element-wise equality stress constraints to a magnitude close to the yield strength of the material. This computation generated the initial solid spar thickness and subsequently, the size of void within each element in the grid, using mass function models derived from a combination of Tsai-Hill failure Theory and Beam Theory. In the second step, Finite Element Analysis in a high-resolution grid was executed while regenerating webs within the voids in the spar model to impose stiffness at mass penalty. A Pareto optimal curve, derived from the mass and deflection response of the model at each increment of the input variable was subjected to the Utopian point optimality criteria in order to determine the optimum combination of mass and stiffness response. The resulting wing spar showed a 30.7% reduction in mass and 45.7% reduction in the maximum deflection when compared to the baseline model. The model demonstrates more wing spar mass savings using a Fully Stressed Element Topology approach, compared to the mass of the baseline spar that was obtained using Taguchi based sizing optimization approach.

Key words: Regeneration, Stress constraint, Fully Stressed, Pareto, Optimization, Unmanned Aerial Vehicle, Topology, Multiresolution