Real-time snow simulation with compression and thermodynamics
Abstract: Background: Snow simulation can be used to increase the visual experience in applications such as games. Previously, snow has been simulated in real-time through two-dimensional grid based methods, which limits itself in the aspect of dynamic interactions. To widen the scope of what games current game engines can produce, an approach to simulating the behavior of snow with non-recoverable compression and phase transition is proposed. Objective: The objective of this thesis is to construct a particle simulation model to simulate the behaviors of snow in regards to compression and phase transition in real-time. The solution is limited to the behavior of deposited snow, and will therefore not consider the aspect of snowfall and realistic visualization. Method: The method consists of a particle simulation with incorporated functionality of compression and thermodynamics. Test cases based on compression, phase transition and performance have been conducted. Results and Conclusions: The results show that the model captures phase transition with the phases of snow, ice, and water. Compression by external forces and self-weight is also captured, but with missing behavior in terms of bond creation between grains. Performance measurements indicates that the simulation is applicable for real-time applications. It is concluded that the approach is a good stepping stone for future improvements of snow simulation.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)