PARTICLES 2023

Fast Point-To-Mesh Distance Computation Technique Based On Cell Linked List For Polygon-Wall Boundary In Moving Particle Semi-Implicit Method

  • Teixeira, Matheus (University of São Paulo)
  • Pereira, Lucas (University of São Paulo)
  • Amaro Jr, Rubens (University of São Paulo)
  • Cheng, Liang-Yee (University of São Paulo)

Please login to view abstract download link

The simulation of fluid-structure interaction (FSI) problems usually involves many degrees of freedom, and a considerable number of particles is generally required to model both fluid and solid domains. In relation to the modeling of solid walls by particles, the use of triangular meshes provides more efficient and smoother representation of complex-shaped solid surfaces as well as the straight coupling between particle and mesh-based methods, which is suitable for FSI applications. However, in the particle-based simulations with solid boundaries modeled by mesh, the computation of the particle-mesh distances is a critical time-consuming task, and a fast technique is of major importance. Taking advantage of the cell linked list structure widely adopted for fixed-radius neighborhood search algorithms in particle methods, we proposed a Fast Point-to-mesh Distance computation technique based on Cell linked list (FPDC). Alongside this new technique, a particle-polygon wall contact model was introduced to enable simulations of the collision between the surface of the moving bodies and fixed wall represented, respectively, by particles and mesh. The results show that the proposed technique provides a significant processing time speedup and can be used for practical large-scale problems.