检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:朱昶胜[1,2] 金显 邓新[1] 冯力[2] ZHU Chang-sheng;JIN Xian;DENG Xin;FENG Li(College of Computer and Communication,Lanzhou Univ.of Tech.,Lanzhou 730050,China;State Key Laboratory of Advanced Processing and Recycling of Nonferrous Metals,Lanzhou Univ.of Tech.,Lanzhou 730050,China)
机构地区:[1]兰州理工大学计算机与通信学院,甘肃兰州730050 [2]兰州理工大学省部共建有色金属先进加工与再利用国家重点实验室,甘肃兰州730050
出 处:《兰州理工大学学报》2018年第2期21-27,共7页Journal of Lanzhou University of Technology
基 金:国家自然科学基金(51661020;11504149;11364024);中国博士后基金(2014M560371)
摘 要:材料微观组织数值模型模拟是一个密集型计算问题,其模拟时间太长且模拟规模太小.特别是在反映现实模拟的三维多场耦合材料枝晶成型过程中,由于模拟规模太小和计算时间太长,从而导致不能清楚地、及时地反映出枝晶的生长过程.为解决这两个问题,提出使用MPI对等模式对耦合流场的相场法进行三维晶枝生长模型模拟计算,并沿x轴等值面切割整个模型,把分割后的小模型分到不同MPI节点中实现并行运算.结果表明:在相同模拟规模下,10个MPI并行计算节点的加速比可达串行的19.9倍;同时其模拟规模也从串行的211×211×211个网格数增加到388×388×388个网格数.证明使用MPI并行计算对PF-LBM进行模拟解决了单CPU上模拟规模太小和计算时间太长的问题.Numerical model simulation of material macrostructure is an intensive computational problem where the simulation time is too long and the simulation scope is too small.Especially,in the formation process of dendrites that reflects realistic simulation of three-dimensional multi-field coupling of material,the growth process of dendrites can not be clearly and promptly reflected because the simulation scope is too small and the calculation time is too long.To solve these two problems,the phase-field method of coupled flow field with MPI reciprocal mode is proposed to compute simulative the three-dimensional crystal growth model and the entire model is cut along isoparametric surface in x-axis direction,and the small models after cutting are distributed onto different MPI nodes to realize the parallel operation.The result shows that in the same simulation scope,the acceleration ratio of 10 MPI parallel computation nodes will amount to 19.9 times of that of serial computation nodes.Meantime,the simulation scope will increase from serial grid number of 211×211×211 to that of 388×388×388.It is proved that the application of MPI parallel calculation to the PF-LBM simulation will resolve the problem that the simulation scope is too small and the calculation time is too long on a single CPU.
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.7