Applied Computational Geometry for Smoothed Particle Hydrodynamics

Abstract

Smoothed Particle Hydrodynamics (SPH) is a Computational Fluid Dynamics method that has significantly increased in capability in recent years. The advent of general-purpose computing on graphics processing units (GPGPU) has helped to enable large-scale SPH simulation outside of supercomputer environments. However variable resolution methods for free-surface and multiphase SPH simulations remains limited. The lack of robust and efficient variable resolution methods for free-surface and multiphase simulations limits the potential computational efficiency and accuracy of SPH when used to model these flows. Recent work has made steps towards solving these issues and has improved the capability of SPH in these areas. This thesis seeks to add to these growing capabilities by presenting new computational geometry algorithms to improve the accuracy and efficiency of SPH simulations. These methods include a boundary condition for flexible walls, a spatial filter to reduce particle disorder, and a modified Voronoi tessellation method for determining particle volumes. These methods provide a basis for future work to increase the computational efficiency of SPH and will facilitate further improvements to the accuracy and scalability of SPH simulations involving free-surface and multiphase flows.