A Distance Regularization Scheme for Topology Optimization with Parametric Level Sets Using Cut Elements

Lin Yi,Zhu Weidong,Li Jiangxiong,Ke Yinglin
DOI: https://doi.org/10.1007/s00158-021-03098-5
IF: 4.279
2022-01-01
Structural and Multidisciplinary Optimization
Abstract:The regularization of the level set function is important to ensure numerical stability in the level set topology optimization. For the parametric level set method, introducing the distance potential functional is a popular way to regularize the distance of level set field and bypass the common re-initialization procedure in the conventional level set method. However, the conventional distance potential functional does not have sufficient control on the gradient of level set function and therefore it may lead to the lack of smoothness and regularity for the optimization results. This paper presents a modified potential functional which incorporates the information of the interfaces into the domain of functions so that the signed distance property can be enforced directly along the interfaces. In addition, the diffusion term is employed to further smooth the structural boundaries. The Compactly Supported Radial Basis Functions (CSRBF) are used to parametrize the level set function and the cut element method is adopted to discretize the governing equations. The linear elasticity problems including compliance minimization problems and compliant mechanism problem are tested to validate the proposed method.
What problem does this paper attempt to address?