Physically-based Process Of Generating And Rendering Technology For High Energy Explosion

With the development of science technology, Visual scene simulation has infiltrated every aspect of military affairs. Macroscopic vision of aerosphere high energy explosion is one of the important foundations for effect estimating of the explosion intuitively. The appearance based model can not suffice to demand of experiment and drill, physically-based macroscopic vision simulation of high energy explosion has become a hot topic of computer graphics.Based on deeply study of physically-based fluid simulation and rendering methods, we have analyzed the comparisons of simulation methods and adoption of rendering methods according to the categories and representations of fluid models. We use Euler grid based method and Lagrangian particle based method to create mushroom cloud and shock wave. The contributions and relevant work in the paper are as follows.Firstly, a Lagrangian particle based simulation method is presented to animate shock wave. By creating force model of shock wave smacking the ground, we use Smoothed Particle Hydrodynamics method to simulate the sand flow. An eight grids algorithm is presented for space search. Through dividing grid by three-dimensionality, we capture the main area of neighbor particles. Efficiency of search will enhance 70.4% compared with neighbor grid space search. The capturing area of neighbor particles will enhance 46.4% compared with traditional eight grids algorithm.Secondly, different explosion models are presented based on different force structure. The form of previous simulation of mushroom cloud is different from the real scene. We design force structure according to different condition, and it has good adaptability and represents the structural character of mushroom figure.Thirdly, we proposed the semi-Lagrangian based macroscopic vision effect of high energy explosion. Vorticity confinement force based on each point is adopted to reduce numerical dissipation of semi-Lagrangian advection. Catmull-Rom spline is adopted to enhance accuracy of scalar computing, and to reduce sharp edge of model.Fourthly, to avoid the traditional semi-Lagrangian method which tapers the simulation of smoke, buoyancy equation is designed from density and temperature via solution of advection and diffusion equation. The simulation result is rendered with Planck's radiation law based on temperature in each point.Finally, we use GPU parallel computation technology to realize real-time simulation by synthesizing the algorithms above. We enlarge grid size about 8~10 times larger than the traditional one, and enhance the accuracy of computation and rendering.