Abstract
Dynamic boundary condition (DBC) has been widely used in SPH method. However, in certain situations, it was found that a few fluid particles could break through the boundary or were not reflected specularly. Of course, these phenomena are unphysical. To improve the performance of DBC, an improved dynamic boundary condition (IDBC) was presented in this paper. To prevent fluid particles from breaking through the boundary, the repulsive force of boundary particles was enhanced by expanding the equation of state into a higher order. To deal with the asymmetry of DBC, a rectangular support domain attached to boundary particles and a corresponding correction factor are proposed. The results of three test cases showed that the performance of IDBC was satisfied.
Highlights
Smoothed particle hydrodynamics (SPH) was initially proposed by Monaghan and Gingold [1], and by Lucy [2] in 1977
This paper presents an improved dynamic boundary condition (IDBC) in SPH methods with three improvements
The fluid particle cannot return the initial position after only one loop in Dynamic boundary condition (DBC), and the energy fluctuates unphysically
Summary
Smoothed particle hydrodynamics (SPH) was initially proposed by Monaghan and Gingold [1], and by Lucy [2] in 1977. SPH has progressed tremendously due to intensive theoretical work and computational improvements, since the 1990s This numerical method has been shown to be robust and applicable to a wide variety of fields. Different types of boundary conditions can be used, such as ghost particles [6], repulsive particles [3], LUST [7], and so on. The differences are that the boundary particles are not allowed to move, or they can only move according to some external input and the density of boundary particles are kept constant In this way, all the variables associated with the particles can be calculated in the same manner with a considerable saving of computation time. In some simulations, we found a few fluid particles broke through the boundary This phenomenon is unphysical, and it possibly indicated that the repulsive force was not strong enough. The original kernel is retained, and the algorithm keeps concise
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have