PSI - Issue 79

Available online at www.sciencedirect.com

ScienceDirect

Procedia Structural Integrity 79 (2026) 394–403

© 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of IGF28 - MedFract3 organizers Keywords: Deep learning; Anisotropy; Voronoi tessellation; Finite element; Trabecular bone Abstract Accurate prediction of bone fragility requires understanding how the anisotropic and disordered trabecular architecture governs nonlinear mechanical behavior. However, hierarchical organization, intrinsic disorder, and material heterogeneity hinder both direct numerical modeling and the development of robust deep learning surrogates. To address these challenges, we initiate the development of a general deep learning framework for predicting the mechanical response of anisotropic, disordered architectures using simplified yet representative Voronoi-based structures that emulate trabecular networks. A dataset of 10 000 two-dimensional beam-based Voronoi tessellations with tunable anisotropy and cell density was generated and analyzed through nonlinear finite element simulations under uniaxial compression with periodic boundary conditions. Nonlinear structural responses were quantified via e ff ective sti ff ness and strength, examining trends with respect to geometric descriptors. Three Multi-Layer Perceptron models were trained using di ff erent input representations, from global descriptors to explicit nodal coordinates. The model trained on global parameters achieved the best predictive accuracy ( R 2 = 0 . 77for sti ff ness, R 2 = 0 . 88 for strength), indicating that coarse geometrical descriptors are su ffi cient to capture first-order mechanical trends, while coordinate-based inputs highlight the limitations of MLP architectures for variable-size structural data. These results demonstrate that simple models can predict key e ff ective properties from global geometry, establishing a foundation for data-driven modeling of trabecular analogs. Future work will integrate graph-based architectures, enhanced datasets, and 3D nonlinear simulations to develop a unified predictive framework for bone-inspired and architected materials. 28th International Conference on Fracture and Structural Integrity - 3rd Mediterranean Conference on Fracture and Structural Integrity Anisotropy and failure in trabecular structures: a deep learning approach Marco Piacentini a , Chiara Bertolin a , Bjørn Skallerud b , Filippo Berto c , ChaoGao a, ∗ a Norwegian University of Science and Technology (NTNU), Department of Mechanical and Industrial Engineering, Richard Birkelands vei 2B, 7491 Trondheim, Norway b NTNU, Department of Structural Engineering, Richard Birkelands vei 1A, 7034 Trondheim, Norway c Sapienza University, Department of Chemical Engineering Materials Environment, Via Eudossiana 18, 00184 Roma, Italy

∗ Corresponding author. E-mail address: chao.gao@ntnu.no

2452-3216 © 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of IGF28 - MedFract3 organizers 10.1016/j.prostr.2025.12.350