Lumping Techniques for Mixed Finite Element Diffusion Discretizations

被引:3
作者
Maginot, P. G. [1 ]
Brunner, T. A. [1 ]
机构
[1] Lawrence Livermore Natl Lab, 7000 East Ave, Livermore, CA 94551 USA
关键词
Diffusion; high-order; lumping; mixed finite element; positivity; TRANSPORT; SOLVER;
D O I
10.1080/23324309.2018.1461653
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We are interested in positivity preserving spatial discretizations applicable to arbitrary order mixed finite element spatial discretizations of the radiative diffusion equations. Though not guaranteed to yield strict positivity, matrix lumping is appealing as a method of increasing solution positivity because unlike ad hoc fixups or strictly non-negative, non-linear solution representations, matrix lumping keeps the mixed finite element diffusion equations as a linear system of equations. Self-lumping schemes have previously been shown to be a generalization of more traditional matrix lumping techniques. Self-lumping schemes restrict integration quadrature points to the interpolation points used to define interpolatory finite element basis functions. As shown previously, the positivity and order of accuracy of self-lumping schemes are directly related to the choice of finite element interpolation point. In this work, we develop self-lumping schemes for arbitrary order, mixed finite element spatial discretizations of the diffusion equation in Cartesian geometry. After introducing several possible self-lumping schemes appropriate for mixed finite elements, we test the ability of each scheme to increase solution positivity relative to an unlumped discretization using a simple test problem designed to induce negative solutions, evaluating efficacy as a function of finite element order and finite element interpolation point type. The method of manufactured solutions is then used to evaluate order of convergence for promising self-lumping schemes.
引用
收藏
页码:301 / 325
页数:25
相关论文
共 50 条
[21]   A mixed finite element solver for natural convection in porous media using automated solution techniques [J].
Zhang, Chao ;
Zarrouk, Sadiq J. ;
Archer, Rosalind .
COMPUTERS & GEOSCIENCES, 2016, 96 :181-192
[22]   Finite element microstructural homogenization techniques and intergranular, intragranular microstructural effects on effective diffusion coefficient of heterogeneous polycrystalline composite media [J].
Jothi, S. ;
Croft, T. N. ;
Brown, S. G. R. ;
de Souza Neto, E. A. .
COMPOSITE STRUCTURES, 2014, 108 :555-564
[23]   Error analysis for discretizations of parabolic problems using continuous finite elements in time and mixed finite elements in space [J].
Bause, Markus ;
Radu, Florin A. ;
Koecher, Uwe .
NUMERISCHE MATHEMATIK, 2017, 137 (04) :773-818
[24]   Convergence of an upwind control-volume mixed finite element method for convection-diffusion problems [J].
Rui, H. .
COMPUTING, 2007, 81 (04) :297-315
[25]   Finite element simulation of diffusion into polycrystalline materials [J].
Gryaznov, D. ;
Fleig, J. ;
Maier, J. .
SOLID STATE SCIENCES, 2008, 10 (06) :754-760
[26]   Finite Element Analysis of Hydrogen Diffusion in Materials [J].
Kanayama, Hiroshi ;
Ndong-Mefane, Stephane ;
Ogino, Masao .
EFFECTS OF HYDROGEN ON MATERIALS, 2009, :580-+
[27]   Mixed finite element formulation for folded plates [J].
Eratli, N ;
Aköz, AY .
STRUCTURAL ENGINEERING AND MECHANICS, 2002, 13 (02) :155-170
[28]   A multipoint flux mixed finite element method [J].
Wheeler, Mary F. ;
Yotov, Ivan .
SIAM JOURNAL ON NUMERICAL ANALYSIS, 2006, 44 (05) :2082-2106
[29]   A multiscale mortar mixed finite element method [J].
Arbogast, Todd ;
Pencheva, Gergina ;
Wheeler, Mary F. ;
Yotov, Ivan .
MULTISCALE MODELING & SIMULATION, 2007, 6 (01) :319-346
[30]   A Mixed Finite Element Method for Coupled Plates [J].
Hu, Jun ;
Liu, Zhen ;
Ma, Rui ;
Wang, Ruishu .
COMPUTATIONAL METHODS IN APPLIED MATHEMATICS, 2025, 25 (03) :601-618