Flow2Mesh: A flow-guided data-driven mesh adaptation framework

被引:0
|
作者
Yu, Jian [1 ]
Lyu, Hongqiang [2 ]
Xu, Ran [1 ]
Ouyang, Wenxuan [2 ]
Liu, Xuejun [1 ]
机构
[1] Nanjing Univ Aeronaut & Astronaut, Coll Comp Sci & Technol, Minist Ind & Informat Technol, Key Lab Pattern Anal & Machine Intelligence, Nanjing 211106, Peoples R China
[2] Nanjing Univ Aeronaut & Astronaut, Coll Aerosp Engn, Nanjing 211106, Peoples R China
关键词
FINITE-ELEMENT METHODS;
D O I
10.1063/5.0188690
中图分类号
O3 [力学];
学科分类号
08 ; 0801 ;
摘要
Mesh adaptation is crucial in numerical simulation, providing optimal resource allocation for accurately capturing physical phenomena. However, when applied to Computational Fluid Dynamics (CFD) problems with complex multi-scale properties, existing adaptation methods face huge challenges due to the high computational cost of solving auxiliary partial differential equations (PDEs) and the difficulty in aligning the flow features with mesh geometric features. In this work, an end-to-end data-driven mesh adaptation framework, Flow2Mesh, is proposed to address these challenges by adopting a hybrid modeling strategy to construct the mapping from pixelated flow-fields to graph-based meshes. It achieves a rapid and accurate one-step mesh adaptation via a perceptual feature network (PFN) and a mesh movement network (MMN). PFN extracts the global perceptual features from flow-fields to enhance flow feature representation and mesh resolution independence. In MMN, these features are utilized to deform the initial mesh to a topology-invariant adaptive mesh by a proposed physically driven mesh convolutional network. It considers the inherent mesh geometric information for efficient node feature aggregation and alignment of mesh density with a flow-field structure. To generate high-quality adaptive meshes, various mesh-related losses are designed to regularize the mesh movement and alleviate the mesh tangling. Experiments in CFD scenarios demonstrate the generalization of our model to different design parameters and mesh configurations. It takes three orders of magnitude less time to generate similar meshes than the PDE-based method. The results exhibit the potential of Flow2Mesh to be a flexible and reliable tool for rapid mesh adaptation in scientific and industrial fields.
引用
下载
收藏
页数:20
相关论文
共 50 条
  • [1] A FRAMEWORK FOR DATA-DRIVEN PROGRESSIVE MESH COMPRESSION
    Cirio, Gabriel
    Lavoue, Guillaume
    Dupont, Florent
    GRAPP 2010: PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON COMPUTER GRAPHICS THEORY AND APPLICATIONS, 2010, : 5 - 12
  • [2] Data-driven tetrahedral mesh subdivision
    Rodriguez, L.
    Navazo, I.
    Vinacua, A.
    COMPUTER GRAPHICS FORUM, 2007, 26 (04) : 783 - 793
  • [3] Unsteady flow simulation with mesh adaptation
    Tang, Jing
    Zhang, Jian
    Li, Bin
    Zhou, Nai-Chun
    INTERNATIONAL JOURNAL OF MODERN PHYSICS B, 2020, 34 (14-16):
  • [4] Mesh Adaptation for Stationary Flow Control
    R. Becker
    Journal of Mathematical Fluid Mechanics, 2001, 3 : 317 - 341
  • [5] Mesh Adaptation for Stationary Flow Control
    Becker, Roland
    JOURNAL OF MATHEMATICAL FLUID MECHANICS, 2001, 3 (04) : 317 - 341
  • [6] Mesh ensemble motion graphs: Data-driven mesh animation with constraints
    James, Doug L.
    Twigg, Christopher D.
    Cove, Andrew
    Wang, Robert Y.
    ACM TRANSACTIONS ON GRAPHICS, 2007, 26 (04):
  • [7] Error Compensation Framework for Flow-Guided Video Inpainting
    Kang, Jaeyeon
    Oh, Seoung Wug
    Kim, Seon Joo
    COMPUTER VISION - ECCV 2022, PT XV, 2022, 13675 : 375 - 390
  • [8] Mesh adaptation strategies for shallow water flow
    Marrocu, M
    Ambrosi, D
    INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS, 1999, 31 (02) : 497 - 512
  • [9] Code Comments Generation with Data Flow-Guided Transformer
    Zhou, Wen
    Wu, Junhua
    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, 13579 LNCS : 168 - 180
  • [10] Mesh Parametrization Driven by Unit Normal Flow
    Zhao, Hui
    Su, Kehua
    Li, Chenchen
    Zhang, Boyu
    Yang, Lei
    Lei, Na
    Wang, Xiaoling
    Gortler, Steven J.
    Gu, Xianfeng
    COMPUTER GRAPHICS FORUM, 2020, 39 (01) : 34 - 49