基于物理的流体动画是计算机图形学研究的热点之一,然而不管是欧拉法、SPH方法或LBM方法,这些基于物理的方法意味着繁重的计算量,均需要求解复杂的非线性运动方程.由于这些方法都有各自的优缺点,一方面对细节处理的能力不同,另一方面计算耗费有差别,最近几年许多研究者开始探索混合表达的思路,力图对效率和细节进行权衡.因此从目前流体动画的研究方法入手,重点讨论了各类方法的混合使用.提出了不同的混合表达的思路,并指出了混合表达的难点在于各类方法在不同方法计算界面上的一致性问题求解,只有解决好一致性问题才能避免计算错误.未来这方面的研究将集中在多种方法的混合使用上,而且重点在于如何保证守恒量在融合过程中得到满足.
The physically based fluid animation is one of the hottest topics in computer graphics. However, no matter Euler method, SPH or LBM method are time consuming since they always need to solve the complex nonlinear equations. Because each method holds its own advantages and disadvantages on fluid details catching or performance, in recent years many researchers turn to hybrid methods to balance the efficiency and detail expression. This paper tries to summarize the available different hybrid methods. As we point out that for different forms within the fluid simulation, different strategies should be used for some very specific purpose. And the difficulty of hybrid expression is how to maintain the consistency across the interface of these different methods to avoid computation problems. Further research will focus on the fusion of different methods with conservation satisfied across the interface.