重构核粒子法对光滑粒子法的改进效果
Reproducing Kernel Particle Method in Smoothed Particle Hydrodynamics
-
摘要: 光滑粒子法通过核函数进行近似估计,在计算域边界附近,核估计的精度明显下降.重构核粒子法通过校正函数对核函数进行重新构造,提高核估计方法在边界点和内点上对函数的估计精度以及计算的稳定性.研究发现,虽然校正函数的构造立足于对函数的精确估计,但这个优势同样能在对导数的估计中继续保持.通过理论研究及数学、物理模型的模拟,展示重构核粒子法的改进效果,揭示其能够提高精度的原因.Abstract: Kernel functions approximate physical quantities in smoothed particle hydrodynamics (SPH) method.However,approximation accuracy decreases near boundaries.By reconstruction of a kernel with correcting function,reproducing kernel particle method (RKPM) improves approximation accuracy and computational stability at internal or boundary points.Though correction is proposed to approximate functions accurately,it also benefits in approximation of derivatives.