首页> 外文会议>International Conference on Information Science and Applications >Parallel Simulation of an Unsteady Forced Convection around a Linear Cascade Using PC-Cluster
【24h】

Parallel Simulation of an Unsteady Forced Convection around a Linear Cascade Using PC-Cluster

机译:使用PC-Cluster并行模拟线性级联周围的非恒定强迫对流

获取原文

摘要

A multi-core PC cluster based on linux operating system was designed and built with an affordable PC hardware and high speed Gigabyte network switches. The PC cluster consisted of 36-cores and easily expandable. With a commercial CFD (Computational Fluid Dynamics) package, the performance of the HPC was evaluated by changing the number of CPU involved in the computation. An unsteady forced convection problem around a linear cascade was solved using the CFX program and the heat transfer coefficient along the surface of the cascade was investigated. The mesh of the model CFD problem has 1.5 million nodes and the unsteady computation was performed during 2000 time-integrations. The performance of present HPC cluster increases as the number of code increases up to 16 cores. It was shown that the computation time with 16-core is approximately 3times faster than that with 4-core.
机译:设计并构建了基于linux操作系统的多核PC群集,并使用价格合理的PC硬件和高速技嘉网络交换机进行了构建。 PC集群由36核组成,并且易于扩展。使用商业CFD(计算流体动力学)软件包,可以通过更改计算中涉及的CPU数量来评估HPC的性能。使用CFX程序解决了线性叶栅周围的非稳态强迫对流问题,并研究了沿叶栅表面的传热系数。模型CFD问题的网格具有150万个节点,并且在2000次时间积分过程中执行了非定常计算。当前的HPC集群的性能随着代码数量增加到多达16个内核而提高。结果表明,使用16核的计算时间比使用4核的计算时间快约3倍。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号