Divergence-Free Smoothed Particle Hydrodynamics

Jan Bender, Dan Koschier
ACM SIGGRAPH / EUROGRAPHICS Symposium on Computer Animation

In this paper we introduce an efficient and stable implicit SPH method for the physically-based simulation of incompressible fluids. In the area of computer graphics the most efficient SPH approaches focus solely on the correction of the density error to prevent volume compression. However, the continuity equation for incompressible flow also demands a divergence-free velocity field which is neglected by most methods. Although a few methods consider velocity divergence, they are either slow or have a perceivable density fluctuation.

Our novel method uses an efficient combination of two pressure solvers which enforce low volume compression (below 0.01%) and a divergence-free velocity field. This can be seen as enforcing incompressibility both on position level and velocity level. The first part is essential for realistic physical behavior while the divergence-free state increases the stability significantly and reduces the number of solver iterations. Moreover, it allows larger time steps which yields a considerable performance gain since particle neighborhoods have to be updated less frequently. Therefore, our divergence-free SPH (DFSPH) approach is significantly faster and more stable than current state-of-the-art SPH methods for incompressible fluids. We demonstrate this in simulations with millions of fast moving particles.

» Show BibTeX

author = {Jan Bender and Dan Koschier},
title = {Divergence-Free Smoothed Particle Hydrodynamics},
booktitle = {Proceedings of the 2015 ACM SIGGRAPH/Eurographics Symposium on Computer
year = {2015},
publisher = {ACM},
doi = {http://dx.doi.org/10.1145/2786784.2786796}

Disclaimer Home Visual Computing institute RWTH Aachen University