Comparison of Numerical Accuracy between MPS Method and LSMPS Method as Particle Methods in Computational Fluid Simulation
Comparison of Numerical Accuracy between MPS Method and LSMPS Method as Particle Methods in Computational Fluid Simulation
复制标题
计算流体模拟中粒子法MPS法与LSMPS法的数值精度比较
DOI：
10.11540/bjsiam.26.2_2
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
室谷浩平，玉井佑，越塚誠一
中科院分区：
文献类型：
--
作者：
Fangyuan Hu;Takuya Matsunaga;Tasuku Tamai;Seiichi Koshizuka;室谷浩平，玉井佑，越塚誠一
本文通过数值验证比较了移动粒子模拟（MPS）方法和最小二乘移动粒子半隐式（LSMPS）方法。MPS方法是由Koshizuka和Oka于1996年发展起来的，用于求解具有自由表面的不可压缩流动。虽然原始的MPS方法很难准确和稳定地解决各种工业问题，但今天的MPS方法通过各种数值稳定技术，成为工程领域最流行的粒子方法之一。然而，由于MPS方法的空间离散模型不具有一致性，因此MPS方法无法在数学上精确求解微分方程。为了解决这个问题，LSMPS方法由Tamai和Koshizuka于2014年开发。LSMPS方法具有任意高精度的无网格空间离散格式、一致的时间积分格式和边界条件的广义处理。本文在简要介绍MPS方法和LSMPS方法的基础上，对这两种方法进行了两种数值验证。
In this paper, the moving particle simulation (MPS) method and the least squares moving particle semi-implicit (LSMPS) method are compared by numerical verification. The MPS method was developed in 1996 by Koshizuka and Oka in order to solve incompressible flow with free surface. Although the original MPS method was difficult to solve various industrial problems with accuracy and stability, the todayʼs MPS method becomes one of the most popular particle method in the engineering fields by various numerical stabilization techniques. However, since spatial discretization models of the MPS method do not have consistency, the MPS method can not solve mathematically precisely differential equations. In order to solve this problem, the LSMPS method was developed in 2014 by Tamai and Koshizuka. The LSMPS method have arbitrary high-order accurate meshfree spatial discretization schemes, consistent time integration schemes, and generalized treatment of boundary conditions. In this paper, after summaries of the MPS method and the LSMPS method are described, two kinds of numerical verifications are performed for both methods.