A generic smoothed wall boundary in multi-resolution particle method for fluid-structure interaction problem

Dealing with complex geometry boundaries is a challenge for particle boundary conditions. Since the effective area of the fluid particle is cut off by the boundary, it is difficult to obtain accurate results near the boundary, especially for complex geometry shapes. Things become more complicated when dealing with multiple sizes of particles in multi-resolution particle methods. In this paper, a generic smoothed wall (GSW) boundary is proposed by using single layer particles to accurately discretize the boundary. The boundary weight function is transformed to a distance function to fill the vacancy of the particle number density of fluid particles. The GSW boundary can be applied in the static, moving and non-linear deforming wall conditions. It also can be used to effectively handle complex geometric boundaries, since simulation results are not affected by changes in the number and size of wall particles. In addition, the GSW boundary also shows good results when combined with multi-resolution methods solving problems with multiple sizes of particles. In this paper, the MPS-DEM coupling method with GSW boundary model is applied to solve non-linear FSI problem. Several cases have been calculated to validate the accuracy, convergency and stability of the GSW boundary model. Compared with dummy particle boundary, results show that the proposed GSW boundary model can improve the numerical accuracy near the boundary, and handle arbitrary moving boundaries with high efficiency. (c) 2021 Elsevier B.V. All rights reserved.

Automated summary

In this paper, a generic smoothed wall (GSW) boundary model is proposed to accurately handle complex geometric boundaries and multiple particle size problems. By using single layer particles and transforming the boundary weight function into a distance function, the GSW boundary model fills the vacancy of the particle number density of fluid particles. The GSW boundary model can be effectively applied in static, moving, and non-linear deforming wall conditions, and performs well when combined with multi-resolution methods.