Hybrid modeling of Lagrangian–Eulerian method for high-speed fluid simulation

This paper proposes a hybrid Lagrangian–Eulerian method involving all participating media for high-speed fluid simulation and its accompanying phenomena. We take advantages of both particle- and grid-based solvers in our new approach by dividing all the physical states into two different parts based on their individual characteristics. Lagrangian-based solver gives rise to liquid, droplet, and foam creation and facilitates the drastic shape distortion and state transition during the multi-physical processes simulation from the perspective of individual particles. At the same time, Eulerian-based solver mainly focuses on the animation of spray, which enables a fog-like atmosphere. Essential to all of the aforementioned functionalities is the state transition among different media considering both physical and geometric features in this paper. The state transition enables to streamline the dis-association and generation of certain physical states. Moreover, to ensure the synchronization of hybrid models, we establish essential interaction and build tight coupling between the particle- and grid-based subsystems. With the comprehensive experimental results, we have manifested that our new modeling approach is both effective and highly modularized, and is capable of producing convincing visual effects for a wider range of graphics and animation tasks.