A node-based self-supporting topology optimization method accelerated by combined approximation reanalysis

doi:10.1016/j.enganabound.2026.106898

A node-based self-supporting topology optimization method accelerated by combined approximation reanalysis

2026-07-01

0

PRE

AI

DOI：10.1016/j.enganabound.2026.106898

Original

Original request for help

Save

Abstract

En 中文

Topology optimization (TO) and additive manufacturing (AM) form a powerful synergy for designing lightweight, high-performance structures. However, integrating overhang constraints for AM often leads to a substantial increase in computational cost. To address this challenge, this paper presents an efficient node-based Self-Supporting topology optimization framework accelerated by Combined Approximation (CA) reanalysis. The proposed framework introduces three primary advancements. First, the Node-based Smoothed Finite Element Method (NS-FEM) is integrated into TO. Utilizing nodes as design variables mitigates mesh dependency and naturally provides explicit spatial coordinates for geometric support evaluation. Second, a generalized nodebased AM-filter featuring an adaptive spatial binning strategy is developed for unstructured meshes, dynamically linking search dimensions to physical AM parameters for rapid support identification. Third, to address the extreme non-linearity introduced by the AM-filter, the optimizer step size is strictly reduced; this specific restriction is uniquely exploited by embedding the CA reanalysis technique, drastically accelerating structural evaluations without sacrificing accuracy. Collectively, these strategies significantly enhance both the applicability and computational efficiency of the structural design process. The robustness and practical engineering value of the proposed framework are comprehensively validated through a diverse series of numerical examples and physical additive manufacturing tests.

Keywords:

Topology optimization

Additive manufacturing

Node-based smoothed finite element method

(NS-FEM)

Overhang constraints

Combined approximation reanalysis

A node-based self-supporting topology optimization method accelerated by combined approximation reanalysis

Abstract

Journal

Organization

Cited Papers