项目名称: 求解对流扩散方程的全离散间断有限元方法
项目编号: No.11271187
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 数理科学和化学
项目作者: 张强
作者单位: 南京大学
项目金额: 50万元
中文摘要: 间断有限元(DG)方法是目前广泛应用于对流扩散问题的数值方法之一。本申请项目计划以一维和二维对流扩散问题为研究对象,开展一系列的全离散DG格式误差分析。具体研究内容包括各种不同时间推进方式与DG空间离散相结合的全离散DG格式,譬如时间方向采用多步法推进,或者半隐半显的Runge-Kutta方法推进等。本项目还计划开展全离散DG格式的超收敛分析和局部分析,以及求解二维对流扩散问题的算子分裂DG算法。我们将重点研究全离散DG格式中各个时间层的边界条件处理技术,以避免数值精度的损失。本项目研究工作具有鲜明的理论价值和应用前景,为斜率限制器和自适应算法的研究提供理论上的保障,促进DG方法的理论研究进展。
中文关键词: 间断有限元;对流扩散问题;全离散方法;误差分析;超收敛
英文摘要: Discontinuous Galerkin (DG) finite element method has been applied widely for the numerical simulations for convection diffusion equations. In this proposal, we will carry out a series of deep studies on the error estimates for many fully discrete DG algorithms, to solve the convection-diffusion equations in one and two space dimensions. Many time-marchings are considered in this proposal, for example, the multilevel algorithm, the semi-implicit Runge-Kutta algorithms and so on. Starting from the global error estimates in energy norm, we will also consider the superconvergence and the local analysis for the fully discrete DG methods. Furthermore, we will consider the operator-splitting DG methods for two-dimensional problems also. The main issue considered in this proposal is the careful treatment on the boundary condition at each stage time level of fully discrete DG methods, in order to avoid the reduction of accurary order. These studies will develop the theory study of DG methods, and help the study in the fields of slope limiters and adaptive implementations.
英文关键词: Discontinuous Galerkin method;Convection diffusion problems;Fully discrete method;Error estimate;Superconvergence