Efficient surface reconstruction for SPH method and its application to simulation of solid-solid contact and fluid-rigid body interaction

Yihua Xiao*, Duping Zhai, Dongdong Jiang, Jianli Shao

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Explicit surface reconstruction is useful for treating challenging boundary-related problems in smoothed particle hydrodynamics (SPH), for example, high-accuracy contact treatment. In this work, an efficient local surface reconstruction method (LSRM) is proposed. It first identifies boundary layer particles and then employs the Delaunay triangulation technique to reconstruct explicit surfaces from the boundary layer particles. The surface reconstruction efficiency of the LSRM is examined by two-dimensional and three-dimensional test cases and compared with that of a global surface reconstruction method (GSRM) proposed previously. Compared with the GSRM, the LSRM shows an increasing advantage in efficiency as the number of particles increases, and its efficiency can be dozens of times higher when the number of particles is large. The LSRM is incorporated into SPH for contact treatment and is used to simulate some typical contact problems. The simulation results show that the LSRM is applicable to solid-solid contact problems involving friction and large deformation and fluid-rigid body interaction problems with complex free surface phenomena. Compared with the GSRM, the LSRM can produce simulation results with similar accuracy and significantly improve the overall efficiency of SPH simulation.

Original languageEnglish
Article number106086
JournalEngineering Analysis with Boundary Elements
Volume171
DOIs
Publication statusPublished - Feb 2025

Keywords

  • Contact
  • Efficiency
  • Smoothed particle hydrodynamics
  • Surface reconstruction

Fingerprint

Dive into the research topics of 'Efficient surface reconstruction for SPH method and its application to simulation of solid-solid contact and fluid-rigid body interaction'. Together they form a unique fingerprint.

Cite this