Parallel Large-scale Geometric Data Processing and Meshing

Project Overview

  • We study partitioning algorithms for large-scale geometric data to support their distributed and parallel processing.
  • We study high-quality structured (quadrilateral) mesh generation for large-scale geometric data.
  • The applications of this research are in large-scale scientific computing.
    • High-quality structured meshes can result in more stable numerical simulations and analysis. Effective data partitioning allows more efficient problem solving on meshes (such as solving complicated partial differential equations) in high-performance computing (HPC) or other parallel environment.
    • We collaborated with coastal modelers from LSU Civil and Environmental Engineering Department and research scientists from LSU Center for Computation and Technology in processing coastal data for hurricane and storm surge simulation.

Project Members



  • W. Yu and X. Li. Regular Mesh Generation for Large-scale Geometric Data in Coastal Modeling. SuperComputing 2014 Poster.
  • W. Yu and X. Li. A Geometry-aware Data Partitioning Algorithm for Parallel Quad Mesh Generation on Large-scale 2D Regions accepted to International Meshing Roundtable (IMR) 2015 [pdf]
  • X. Li, W. Yu, and C. Liu, Geometry-aware Partitioning of Complex Domains for Parallel Quad Meshing accepted to Computer-aided Design (CAD), Vol. 85, pp. 20-33, 2017.
  • C. Liu, W. Yu, Z. Chen, and X. Li, Distributed Poly-square Mapping for Large-scale Semi-Structured Quad Mesh Generation accepted to Computer-aided Design (CAD), (ACM Solid Modeling 2017), Vol. 90, pp. 5-17, 2017.
  • Gallery (Some generated quad meshes)

  • A Geometric Pipe Model

  • West Bay

  • Matagorda Bay

  • Gulf of Mexico


  • We gratefully acknowledge the support of National Science Foundation and Louisiana State Board of Regents on this research.

  • Other:

    Locations of visitors to this page
    Web Page Hit Counter