上海交通大学 海洋工程国家重点实验室； 高新船舶与深海开发装备协同创新中心；船舶海洋与建筑工程学院， 上海 200240
出版日期:2019-06-28发布日期:2019-07-23通讯作者:顾解忡，男，副教授，博士生导师，电话（Tel.）021-34206280; E-mail：xcgu@sjtu.edu.cn.作者简介:张晓慧（1992-），女，陕西省渭南市人，博士生，研究方向为船舶耐波性的数值模拟.基金资助:国家自然科学基金面上项目（51279105，BC0100097）An Explicit Parallel Successive Over-Relaxation Method for Simulation of 2-Dimensional Incompressible Flows
ZHANG Xiaohui,BAI Junli,GU Xiechong,MA NingState Key Laboratory of Ocean Engineering; Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration; School of Naval Architecture, Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
Online:2019-06-28Published:2019-07-23摘要/Abstract
摘要： 提出一种有限体积显式逐次超松弛并行(FV-pSOR)算法，以提高逐次超松弛(SOR)算法求解不可压缩二维流动控制方程组离散所形成的代数方程组的效率.基于区域分解的思想，将计算域分割成4个子域，构造了离散的一般性代数方程组的显式迭代公式并规划了迭代路径；然后，通过数值求解典型二维方腔流，验证了FV-pSOR算法的有效性.结果表明：与SOR算法相比，所提FV-pSOR算法在计算精度相当的前提下的计算效率提高了数倍.
关键词: 分组显式； 逐次超松弛； 并行计算； 有限体积； 区域分解
Abstract: An explicit finite volume parallel successive over-relaxation (FV-pSOR) iterative method is proposed to improve the efficiency of successive over-relaxation (SOR) iterative method in solving the algebraic equations obtained by discretizing the governing equations of 2-dimensional incompressible flows. Based on the domain decomposition technique, computation domain is separated into four sub-domains, and group explicit SOR schemes are constructed for the algebraic equations with general coefficients in flow problems. Furthermore, the iterative route has been discussed in detail. To confirm availability of this algorithm, a series of calculations for a typical lid-driven cavity flow are performed. It demonstrates that comparing to the SOR iterative method, current FV-pSOR iterative method improves the computation efficiency several folds while the accuracy is not degraded.
Key words: group explicit; successive over-relaxation (SOR); parallel computation; finite volume; domain decomposition
PDF全文下载地址:
点我下载PDF
