REGDIFF is proposed, a generative framework that couples voxel representation with latent space regulation and guided diffusion that outperforms voxel-based generative baselines in geometric plausibility, novelty, and diversity and is suggested to be a strong geometry candidate generator for downstream evaluation.
Abstract
Metamaterials are artificially engineered structures whose mechanical and physical behaviors are strongly shaped by geometry rather than composition. Voxel representation provides a unified format for metamaterial geometry generation, as it can express diverse classes such as truss, shell, and porous structures within a single cubic discretization. However, voxel-based generation faces a plausibility-novelty trade-off: staying close to known geometries helps preserve geometric regularities, while moving away from them is necessary for novelty but may produce degenerate geometries. To address this challenge, we propose REGDIFF, a generative framework that couples voxel representation with latent space regulation and guided diffusion. REGDIFF introduces a repel-and-sink (RAS) mechanism to smooth the latent distribution of plausible geometries, and short-range repulsion (SRR) guidance to discourage generation overly close to known samples while maintaining geometric plausibility. We further contribute a voxel-based benchmark covering truss- and shell-type metamaterial geometries, together with an evaluation module for geometric plausibility, novelty, and diversity. Experiments show that REGDIFF outperforms voxel-based generative baselines, achieving +8.9% in geometric plausibility, +46.4% in novelty, and +128.6% in diversity on average across two datasets. These results suggest that REGDIFF is a strong geometry candidate generator for downstream evaluation. Our code is provided at https://github.com/wzhan24/ReGDiff.
: Designing two-dimensional anisotropic mechanical metamaterial unit cells from prescribed effective properties remains a challenging inverse problem, particularly when directional stiffness and material usage need to be controlled simultaneously. In this work, a data-driven conditional diffusion framework is developed...
Hao Sun, Xiao-Hong Ding, Min Xiong et al.· Computers, Materials & C...· 0 citations
The nonlinear and non-unique relationship between unit-cell topology and bandgap properties motivates the development of complementary data-driven approaches for metamaterial inverse design. This work presents a conditional denoising diffusion probabilistic model (DDPM)-based framework for the on-demand generation of t...
Than V. Tran, S. Nanthakumar, Ya-Bin Jin et al.· Machine Learning for Computa...· 0 citations
Architected materials gain stiffness from beams, plates or minimal-surface shells, yet beam lattices lose efficiency to bending and nodal stress concentration and closed-cell lattices are difficult to manufacture and to clear of powder or resin. Open-cell minimal-surface shells print readily, but their zero mean curvat...
Ajith Annavajhula, Gary J. Cheng· Materials Horizons· 0 citations
Entangled granular materials derive exceptional macroscopic rigidity from topological interlocking of their constituents. However, rationally designing particle geometries to maximize this effect remains a formidable challenge due to the immense configuration space and the prohibitive computational cost of discrete e...
Mechanical metamaterials promise unprecedented control over load transfer by tailoring mesoscale architecture, yet most design strategies optimize global stiffness or energy absorption and only indirectly affect force redistribution. Here we introduce a computational–experimental framework to explicitly design load‐d...
A. Detry, Hannes Holey, Raja Zulkarnain et al.· Advanced Functional Material...· 0 citations
Known for his clear and elegant writing style, Bertsekas shaped fields from control and optimization to large-scale computation and artificial intelligence.