A novel adaptive topology optimization method considering unnecessary element removal and progressive mesh refinement

This paper proposes an adaptive topology optimization (TO) approach considering unnecessary element removal and progressive element refinement. A two-stage density filtering for element removal is developed to relax the design space for next iteration, and therefore make a trade-off between solution quality and optimization efficiency. An isolated element detection and deletion is also conducted between element removal and element refinement operation to guarantee the numerical stability. Two 2D numerical examples and one 3D design problems are investigated to demonstrate the effectiveness of the proposed method. Based on the numerical tests, the applicability range of the proposed method and recommended range of element density threshold are provided as well.