4.7 Article

A weak formulation for solving elliptic interface problems without body fitted grid

期刊

JOURNAL OF COMPUTATIONAL PHYSICS
卷 249, 期 -, 页码 80-95

出版社

ACADEMIC PRESS INC ELSEVIER SCIENCE
DOI: 10.1016/j.jcp.2013.04.025

关键词

Elliptic equations; Interface; Non-body fitted mesh; Finite element method; Jump condition

资金

  1. Louisiana Board of Regents RCS [LEQSF (2008-11)-RD-A-18]
  2. National Natural Science Foundation of China [11126134]
  3. NSF Grant [DMS-1115698]
  4. ARO/MURI Grant [W911NF-07-1-0185]
  5. ONR Grant [N00014-11-1-0602]
  6. Science Foundation of China University of Petroleum - Beijing [YJRC-2013-48]
  7. Direct For Mathematical & Physical Scien
  8. Division Of Mathematical Sciences [1317994] Funding Source: National Science Foundation

向作者/读者索取更多资源

A typical elliptic interface problem is casted as piecewise defined elliptic partial differential equations (PDE) in different regions which are coupled together with interface conditions, such as jumps in solution and flux across the interface. In many situations, such as the interface is moving, the challenge is how to solve such a problem accurately, robustly and efficiently without generating a body fitted mesh. The key issue is how to capture complex geometry of the interface and jump conditions across the interface effectively on a fixed mesh while the interface is not aligned with the mesh and the PDE is not valid across the interface. In this work we present a systematic formulation and further study of a second order accurate numerical method proposed in Hou and Liu (2005) [16] for elliptic interface problem. The key idea is to decompose the solution into two parts, a singular part and a regular part. The singular part captures the interface conditions while the regular part belongs to an appropriate space in the whole domain, which can be solved by a standard finite element formulation. In a general setup the two parts are coupled together. We give an explicit study of the construction of the singular part and the discretized system for the regular part. One key advantage of using weak formulation is that one can avoid assuming/using more regularity than necessary of the solution and the interface. We present the numerical algorithm and numerical tests in 3D to demonstrate the accuracy and other properties of our method. Published by Elsevier Inc.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据