报告题目: | 线性和非线性偏微分方程的高精度数值解 |
报 告 人: | 刘 东 |
美国路易斯安娜科技大学 数学和机械工程系 | |
报告时间: | 6月11日(周三)下午15点 |
报告地点: | 九龙湖数学系第一报告厅 |
相关介绍: | 报告摘要: 谱元素法是一种高阶精度有限元方法。它使用雅可比多项式作为基函数和测试函数,对复杂几何区域也能达到十几到二十几阶精度,取决于如何设置。误差可以表现出指数形式的快速收敛,这就是所谓谱的涵义。谱元素法能同时采用缩小有限元大小和提高基函数的阶的两种方法达到高精度。 三维对流扩撒方程和耐维尔斯托克斯方程的谱元素法数值解将被作为两个算例来展示复杂尺寸下应用不同方式提高计算精度。 另一种较简单的高精度方法是隐式紧凑差分法。和常规的有限差法法比,这种方法可以使矩阵的带宽明显变小,误差展现出代数收敛的特性。应用这种方法,波辛奈斯克方程的数值解显示出六阶精度。 当然这些高精度方法也存在不足之处,如何避免和怎样改善等等,这些问题会在随后讨论。 刘东简介: Ph.D. in Applied Mathematics, Brown University, May 2004; Advisor: George E. Karniadakis, Martin R. Maxey M. Sc. in Applied Mathematics, Brown University, May 2000 Advisor: George E. Karniadakis Ph.D. in Computational Engineering Thermophysics, the Institute of Engineering Thermophysics, Chinese Academy of Sciences, July 1998 Advisor: Wei Zao Gu M. Sc. in Computational Thermophysics, Department of Thermal Engineering, University of Science and Technology Beijing, July 1994 Advisor: Zhong Long Gao B. Sc. in Engineering Thermophysics, Southeast University, May 1991 August 2012 - Present: Tenured Associate Professor and Principal Investigator, Mathematics & Statistics and Mechanical Engineering, Louisiana Tech University, Ruston, Louisiana September 2006 � July 2012: Tenure-track Assistant Professor in Mathematics & Statistics and Mechanical Engineering, Louisiana Tech University, Ruston, Louisiana |