Fully-automated topology interpreter for frame structures manufacturable from long products
Year: 2026
Type: article
Abstract: • Robust skeleton simplification while maintaining the initial topology • Straightening of nearly collinear skeleton lines for improved manufacturability • Manufacturing-oriented profile and joint design • Incorporation of symmetry constraints for a reduction of different parts Topology optimization (TO) is a powerful tool used across various disciplines, exemplary to create lightweight structures. But to utilize its full potential, post processing is required to reconstruct TO results as parametric and feature-based computer-aided design (CAD) models, which is a time-consuming task when done manually. While many approaches for an automated interpretation and reconstruction of TO results focus on additive manufacturing technologies, this work aims for conventionally manufactured frame structures made from long products such as rectangular tubes. For this purpose, a fully-automated topology interpreter is presented, which is in particular suitable for low volume fraction TO results that form truss-like structures. The first process steps of skeleton extraction and linearization incorporate state-of-the-art methods. The novelty of this work are especially the subsequent steps of simplification and straightening. Simplification aims to reduce the number of skeleton lines while maintaining the initial topology. Afterwards, straightening tries to make connected lines collinear. Finally, profile cross sections are extruded along the skeleton lines before getting trimmed to generate a parametric CAD model, which is usable for a subsequent size optimization to define the final profile dimensions. Applicability is demonstrated through several examples and sensitivities are investigated qualitatively.
Source: Results in Engineering
Authors: Lukas Kömm, Kristin Paetzold-Byhain
Institution: Technische Universität Dresden
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,940
OpenAlex ID: W7125489270