How to train your solver: Verification of boundary conditions for smoothed particle hydrodynamics
arXiv:2208.10848 · doi:10.1063/5.0126234
Abstract
The weakly compressible smoothed particle hydrodynamics (WCSPH) method has been employed to simulate various physical phenomena involving fluids and solids. Various methods have been proposed to implement the solid wall, inlet/outlet, and other boundary conditions. However, error estimation and the formal rates of convergence for these methods have not been discussed or examined carefully. In this paper, we use the method of manufactured solution (MMS) to verify the convergence properties of a variety of commonly employed of various solid, inlet, and outlet boundary implementations. In order to perform this study, we propose various manufactured solutions for different domains. On the basis of the convergence offered by these methods, we systematically propose a convergent WCSPH scheme along with suitable methods for implementing the boundary conditions. We also demonstrate the accuracy of the proposed scheme by using it to solve the flow past a circular cylinder. Along with other recent developments in the use of adaptive resolution, this paves the way for accurate and efficient simulation of incompressible or weakly-compressible fluid flows using the SPH method.
References in corpus (4)
- Efficient and Accurate Adaptive Resolution for Weakly-Compressible SPH
- Techniques for second order convergent weakly-compressible smoothed particle hydrodynamics schemes without boundaries
- How to train your solver: A method of manufactured solutions for weakly-compressible SPH
- Parallel adaptive weakly-compressible SPH for complex moving geometries
Cited by in corpus (5)
- Robust Solid Boundary Treatment for Compressible Smoothed Particle Hydrodynamics
- Adaptive Compressible Smoothed Particle Hydrodynamics
- Dynamical pressure boundary condition for weakly-compressible smoothed particle hydrodynamics
- Rapid Variable Resolution Particle Initialization for Complex Geometries
- A consistent δ-Plus-ULPH model towards higher accuracy and lower numerical dissipation with fewer neighboring particles