Topology optimization with void exclusion constraint based on a novel virtual physical field method

This is the authors' abstract. We don't add key points for this paper.

Excluding enclosed voids during design remains a critical challenge in additive manufacturing, impacting performance and manufacturability. This study presents a topology optimization method featuring a novel explicit void exclusion constraint. Unlike classical methods requiring additional physical fields, our innovative virtual displacement method addresses common mechanical problems through just one type of physical field, leveraging virtual compliance with polarized distribution for void or solid detection. The void exclusion constraint incorporates a penalized auxiliary volume term, which simplifies parameter tuning, eases computation, and facilitates integration. Reducing degrees of freedom, avoiding global stiffness matrix assembly, and simplifying sensitivity analysis sharply improve the efficiency of finite element analysis for the virtual displacement field. Besides, this approach combines density-based topology evolution for stable iteration and level-set-based result smoothing. The systematic method can eliminate predefined internal or boundary-adjacent voids, regulate local topology, simplify structures, and transform slender or thin-walled designs into thicker ones with minimal performance deterioration. It is implementable, cost-effective, performance-enhanced, efficient, and versatile. It has been validated via 2D/3D examples, 3D-printed prototypes, and comparisons with existing methodologies.