| To reduce product design development time, structural optimization has become an important element within computer-aided-design, for example, the optimal lightweight design of a car door stiffener. Most structural optimization procedures start from a given design topology and then select the shape and size of structural members to achieve optimality of a design objective subject to various constraints. Thus, the success of this approach is highly dependent on a "good" guess of the initial topology.;In this thesis, an integrated structural optimization system is introduced that combines rigorous topology and geometry optimization. This integrated system uses the homogenization method to generate the optimal topology of the initial design and then applies design sensitivity analysis and non-linear programming to obtain the optimal structure.;In the geometry optimization process, design sensitivity analysis is the most computationally expensive procedure. To reduce the cost of product development, faster sensitivity analysis schemes are introduced for elastostatic and eigenvalue optimization problems. Also, problem formulations and design modeling techniques play crucial roles in the optimization process. In this thesis, a detailed discussion of the formulation of structural optimization is provided and a new modeling scheme called the design domain approach is introduced. Using this integrated system approach, results are presented from optimization of several practical two-dimensional and three-dimensional designs. |
