基于Delaunay网格技术的松弛迭代粒子追踪算法

Relaxation based PTV with Delaunay triangulation

  • 摘要: 为提高基于松弛迭代的粒子追踪算法(NRX)的独立性和计算效率,引入DT非结构化网格方法分别代替固定阈值确定与目标粒子保持“拟平行”运动的参选粒子集合以及预匹配的候选粒子集合,并通过模拟二维应变流中单极涡运动对上述尝试进行了全方位的验证。结果表明:一方面,DT方法可以有效地确定参选粒子集合;另一方面,由于DT方法“就近性”的选取特性与确定候选粒子集合的物理背景相悖,不能合理地确定候选粒子集合。在NRX算法的基础上,应用DT方法代替固定阈值确定候选粒子集合,减少一个固定阈值的选取,提出全新的DT-NRX算法;与传统的NRX相比,DT-NRX在高浓度粒子流场分析上更为高效。

     

    Abstract: To improve the independence and matching efficiency of NRX ( New Relaxation based Particle Tracking Algorithm), instead of settled thresholds, DT (Delaunay Triangulation) unstructured grid was introduced to determine the matching candidate particles set and reference particle set in which the selected particles are close to and quasi parallel with the target particle. The validity of introduced DT was thoroughly tested by simulated motion of monopolar vortices in a strain flow. It indicates that DT can determine the reference particle set effectively, while DT fails to determine the candidate particle set because the nearbyselected property of DT is inconsistent with the physical property of particles termed as being neighboring of candidate particles. Finally, on the basis of NRX, a new algorithm DTNRX was proposed by using DT to determine the reference particle set. Compared to the conventional NRX, DTNRX behaves more independently by abandoning one parameter and more effectively in highdensity cases.

     

/

返回文章
返回