A New Tool for Topology Optimization With Gradient-Guided Spectral Level Set Methodology

摘要

In this work we describe and evaluate a new software tool with GUI for structural topology optimization which uses the spectral level set methodology and that can be easily integrated into a multidisciplinary design environment. The spectral level set methodology is a tool to formulate topology optimization problems. Its main advantage is the reduction of the design space dimension under the assumption of a sufficiently smooth interface of interest. This translates into a more efficient way to solve structural topology optimization problems, thus increasing the overall performance of the multidisciplinary optimization strategy. In this paper, we give particular emphasis to known results from functional analysis, in particular the coarea forrmula, and their contribution to the development of the sensitivity analysis of the proposed methodology. We then consider and evaluate strategies to implement the gradients of the functionals involved in the statement of the optimization problem. The proposed approach is integrated into a Windows-based GUI which displays the evolution of the boundary of interest and the final design. This new software tool is then evaluated with a simple topology optimization example with known analytic solution.