SIM-AM 2025

Geometric Parameterization and Analysis of Thin-Walled Additively Manufactured Triply Periodic Minimal Surface Structures

  • Jafari, Abbas (BAM)
  • Chatzi, Eleni (IBK, ETH Zurich)
  • Unger, Jörg F (BAM)

Please login to view abstract download link

Additive manufacturing (AM) has emerged as a transformative technology for fabricating meta-materials with intricate geometries and tunable macroscopic properties, enabling broad applications across fields such as engineering and biomedicine. A well-known challenge in harnessing the full potential of AM is the tendency of the process to introduce geometric imperfections into printed objects. Such deviations can impact the desired mechanical behavior and performance, which is typically predicted using physics-based simulations that assume idealized, nominal geometries. Therefore, analyzing the “as-is” geometry of AM structures is crucial to ensure the reliability of printed components and their simulators [1, 2]. This work centers on the geometric analysis of thin-walled additively manufactured Triply Periodic Minimal Surface (TPMS) structures. We demonstrate the sensitivity of solid-element-based Finite Element (FE) models to geometric imperfections, highlighting the need to incorporate geometric details of as-printed objects into simulations. To manage the extensive geometric variability and deviations inherent in printed specimens, we propose a regularization approach based on two key geometric descriptors, namely “mid-surface position” and “wall thickness”. These features offer two benefits. First, they provide meaningful metrics for quantitatively comparing printed and nominal geometries. Secondly, they enable the construction of shell-based FE models for thin-walled structures, enriched by the features of the actual printed geometry. One advantage of such shell-based FE models is its significantly lower computational cost compared to a solid FE model, enabling a more efficient incorporation of uncertainty within the estimated geometric features. We demonstrate the procedure for extracting and estimating the two geometric parameters for real printed specimens, leveraging computer tomography (CT) imaging and the derived Signed Distance Field. The advantages of integrating these estimated geometric features into FE simulations are demonstrated through a series of analyses based on real printed specimens. REFERENCES: [1] Lozanovski, B., et al. Computational modelling of strut defects in SLM manufactured lattice structures. Materials & Design (2019). [2] Cao, X., et al. Mechanical performance and defect analysis of the imperfect micro smooth gyroid cylinder shell structure. Composite Structures (2021).