A particle-based method for preserving fluid sheets
A particle-based method for preserving fluid sheets
复制标题
DOI：
10.1145/2019406.2019408
复制
发表时间：
2011-08
期刊：
影响因子：
--
通讯作者：
R. Ando;R. Tsuruno
中科院分区：
文献类型：
--
作者：
R. Ando;R. Tsuruno
我们提出了一种新的基于粒子的方法，它显式地保留了用于设置液体动画的薄流体片。我们的主要贡献是一种基于无网格粒子的框架，它在细点分裂，在密集点坍塌，以防止液体分解。与现有的曲面跟踪方法相比，该框架不受数值扩散和纠缠的影响，并且通过无网格表示来稳健地处理拓扑变化。作为基本的流体模型，我们使用流体隐式粒子(Flip)和弱弹簧力来生成平滑的基于粒子的液体动画，该动画在存在涡流或惯性运动的情况下保持均匀的空间粒子分布。通过执行主成分分析(PCA)来检测相邻粒子分布的伸展来检测薄特征，主成分分析用于重建具有各向异性核的薄表面。我们的算法实现直观，易于并行化，能够生成视觉上复杂的稀薄的液体动画。
We present a new particle-based method that explicitly preserves thin fluid sheets for animating liquids. Our primary contribution is a meshless particle-based framework that splits at thin points and collapses at dense points to prevent the breakup of liquid. In contrast to existing surface tracking methods, the proposed framework does not suffer from numerical diffusion or tangles, and robustly handles topology changes by the meshless representation. As the underlying fluid model, we use Fluid-Implicit-Particle (FLIP) with weak spring forces to generate smooth particle-based liquid animation that maintains an even spatial particle distribution in the presence of eddying or inertial motions. The thin features are detected by examining stretches of distributions of neighboring particles by performing Principle Component Analysis (PCA), which is used to reconstruct thin surfaces with anisotropic kernels. Our algorithm is intuitively implemented, easy to parallelize and capable of producing visually complex thin liquid animations.