Generative Lattice Units with 3D Diffusion for Inverse Design: GLU3D

被引:5
作者
Jadhav, Yayati [1 ]
Berthel, Joeseph [1 ]
Hu, Chunshan [1 ]
Panat, Rahul [1 ]
Beuth, Jack [1 ]
Farimani, Amir Barati [1 ]
机构
[1] Carnegie Mellon Univ, Mech Engn, 5000 Forbes Ave, Pittsburgh, PA 15213 USA
关键词
3D diffusion; DDPM; generative model; inverse design; lattice unit cells; TOPOLOGY OPTIMIZATION; EXACT COMPUTATION; METAMATERIALS; COMPOSITES; STIFFNESS;
D O I
10.1002/adfm.202404165
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
Architected materials, exhibiting unique mechanical properties derived from their designs, have seen significant growth due to the design versatility and cost-effectiveness offered by additive manufacturing. While finite element methods accurately evaluate the mechanical response of these structures, identifying new designs exhibiting specific mechanical properties remains challenging, often requiring computationally expensive simulations and design expertise. This underscores the need for a framework that generates structures based on desired mechanical properties without requiring expert input. In this work, a novel denoising diffusion-based model is presented that generates complex lattice unit cell structures based on desired mechanical properties, manufacturable via additive techniques. The proposed framework generates unique lattice unit cell structures in the implicit domain which can be easily converted to mesh structures for fabrication and voxel structures for structural analysis. The proposed model accelerates the design process by generating unique structures with both isotropic and anisotropic stiffness, outperforming traditional unit cells like simple cubic and body-centered-cubic in energy absorption and compression load at lower densities. Additionally, this work explores a new class of hybrid structures, derived by combining multiple configurations of triply periodic minimal surface structures using non-linear transition functions, which may offer equivalent or enhanced strength compared to conventional designs. A novel inverse design method for generating 3D lattice unit cells in the implicit domain using a denoising diffusion probabilistic model is presented. This method accelerates design with tailored mechanical properties, producing high-quality meshes of periodic lattice unit cells for fabrication and analysis. A method for generating hybrid structures that outperform conventional BCC designs in energy absorption and maximum compression load at lower densities is also introduced. image
引用
收藏
页数:15
相关论文
共 103 条
  • [31] Subwavelength Lattice Optics by Evolutionary Design
    Huntington, Mark D.
    Lauhon, Lincoln J.
    Odom, Teri W.
    [J]. NANO LETTERS, 2014, 14 (12) : 7195 - 7200
  • [32] Biomimetic armour design strategies for additive manufacturing: A review
    Islam, Muhammed Kamrul
    Hazell, Paul J.
    Escobedo, Juan P.
    Wang, Hongxu
    [J]. MATERIALS & DESIGN, 2021, 205
  • [33] StressD: 2D Stress estimation using denoising diffusion model
    Jadhav, Yayati
    Berthel, Joseph
    Hu, Chunshan
    Panat, Rahul
    Beuth, Jack
    Farimani, Amir Barati
    [J]. COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING, 2023, 416
  • [34] StressGAN: A Generative Deep Learning Model for Two-Dimensional Stress Distribution Prediction
    Jiang, Haoliang
    Nie, Zhenguo
    Yeo, Roselyn
    Farimani, Amir Barati
    Kara, Levent Burak
    [J]. JOURNAL OF APPLIED MECHANICS-TRANSACTIONS OF THE ASME, 2021, 88 (05):
  • [35] TPMS Designer: A tool for generating and analyzing triply periodic minimal surfaces
    Jones, Alistair
    Leary, Martin
    Bateman, Stuart
    Easton, Mark
    [J]. SOFTWARE IMPACTS, 2021, 10
  • [36] Machine learning: Trends, perspectives, and prospects
    Jordan, M. I.
    Mitchell, T. M.
    [J]. SCIENCE, 2015, 349 (6245) : 255 - 260
  • [37] 3D metamaterials
    Kadic, Muamer
    Milton, Graeme W.
    van Hecke, Martin
    Wegener, Martin
    [J]. NATURE REVIEWS PHYSICS, 2019, 1 (03) : 198 - 210
  • [38] Generating three-dimensional structures from a two-dimensional slice with generative adversarial network-based dimensionality expansion
    Kench, Steve
    Cooper, Samuel J.
    [J]. NATURE MACHINE INTELLIGENCE, 2021, 3 (04) : 299 - 305
  • [39] 3D printed compact heat exchangers with mathematically defined core structures
    Kim, Jiho
    Yoo, Dong-Jin
    [J]. JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING, 2020, 7 (04) : 527 - 550
  • [40] Kingma D.P., 2014, arXiv, DOI 10.48550/arXiv.1412.6980