Smoke Simulation with Two-Scale Vorticity Confinement

Abstract

AbstractEnhancing the details of the smoke rendering is a major direction in virtual reality. One common solution is enhancing turbulence by adding vorticity confinement force. The traditional methods mostly add it on the global grid which would lead to expensive cost. To address this problem, we propose an adaptive two-scale vorticity confinement method. With a new scheme of computational grid, our method can determine the region with rich detail adaptively. First, the Navier-Stokes equations are solved on the coarse grid by the semi-Lagrangian method. Then, we divide the coarse grid using the defined threshold of vorticity. Finally, we obtain the velocity field of coarse grid by sampling from the fine grid, and advect density by it. In addition, we combine the helicity method which can avoid “blow up” when calculating vorticity confinement force. Various experiments validated our method.Keywordsphysically-basedsmoke simulationvorticity confinementturbulence