Particle based snow simulation using dendritic structure

Abstract

We propose a particle based simulation for snow, focusing on the rigid body behavior and fracture effect of the snow. One of the most distinguishable features of the snow, compared to other granular materials or fluids, is a combination of rigid body dynamics and powder-like behavior and fracture effect on the cutting surface. Those features are caused by existence of ice bonds and a microstructure of the snow which is in the shape of dendritic structure. We model the ice bonds with mass-spring model and construct the dendritic structure by using minimum spanning tree. The dendritic structure ensures the rigid body dynamics of snow. Also, the structure has property of fractal pattern so that the snow model using dendritic structure ensures that the cutting surface of snow can show the fracture effect. Our result shows that the dendritic structure ensures the rigid body behavior and generates the fracture effect without modeling the fracture effect explicitly. Our method can be easily applied to existing granular material simulation. We adopt granular material simulation using Smoothed Particle Hydrodynamics(SPH) for base simulation. Our method is easy to implement, since the underlying concept is simple.