Enhanced block-structured quadrilateral mesh generation: integrating cross-field and distance field for optimal domain decomposition

Yuanxing Lv, Beiyan Jia, Yuxiang Yan, Cecil G. Armstrong, Trevor T. Robinson, Liang Sun*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

89 Downloads (Pure)

Abstract

Block-structured meshes are favoured in various computational simulations due to their superior computational efficiency and accuracy. While cross-field methods have demonstrated promising capabilities in generating high-quality quadrilateral meshes, they often face challenges such as non-uniform mesh distribution, limit cycles, and complex block structures. Additionally, the effectiveness of these methods heavily relies on the quality and density of the underlying background mesh. To address these limitations, this paper introduces a novel approach that synergizes medial-axis and cross-field methodologies. Our proposed method diverges from traditional four-sided block decompositions, opting instead for a flexible N-sided subdomain strategy guided by both a cross-field and its corresponding distance field. This innovation not only simplifies the decomposition structure but also lessens the dependency on the background mesh’s density, enhancing the method’s robustness and applicability. The paper details the development and validation of this technique, showcasing its efficiency in handling complex geometries compared to existing methods.

Original languageEnglish
Number of pages16
JournalEngineering With Computers
Early online date28 Nov 2024
DOIs
Publication statusEarly online date - 28 Nov 2024

Publications and Copyright Policy

This work is licensed under Queen’s Research Publications and Copyright Policy.

Keywords

  • domain decomposition
  • quad mesh generation
  • cross-field
  • distance field
  • mesh templates

Fingerprint

Dive into the research topics of 'Enhanced block-structured quadrilateral mesh generation: integrating cross-field and distance field for optimal domain decomposition'. Together they form a unique fingerprint.

Cite this