Robust topology optimization of compliant mechanism using multiobjective optimization method

Abstract

This study evaluates a robust optimum design of a compliant mechanism considering uncertainty of the load direction. For the topology optimization, this study adopts a level set based approach incorporating a fictitious interface energy term. The method is known to have several advantages such as the allowance of topological as well as shape changes, and the ability to qualitatively control the geometrical complexity of the obtained optimum configurations. The design objectives are to maximize the displacement at the gripping position and the robustness under variations of the applied load direction. In addition, the eigenfrequency maximization is considered to acquire the structural stiffness. The design problem is formulated as a multiobjective design problem. Conventionally, the robust design problem has been formulated as a weighted sum of the mean value and the standard deviation of the performance index. Integrating the multiobjective optimization to the level set-based topology optimization method, this study adopts the l_p-norm method that can obtain the Pareto solution even when the Pareto frontier is a non-convex form. Through numerical examples of a compliant mechanism design, validity of the proposed method is verified. Then, the robustness of the obtained optimum configuration is discussed.