SUPERCONVERGENCE OF A DISCONTINUOUS GALERKIN METHOD FOR FRACTIONAL DIFFUSION AND WAVE EQUATIONS

被引:157
作者
Mustapha, Kassem [1 ]
Mclean, William [2 ]
机构
[1] King Fahd Univ Petr & Minerals, Dept Math & Stat, Dhahran 31261, Saudi Arabia
[2] Univ New S Wales, Sch Math & Stat, Sydney, NSW 2052, Australia
关键词
finite elements; dual problem; postprocessing; FINITE-DIFFERENCE METHOD; IMPLICIT NUMERICAL-METHOD; EVOLUTION EQUATION; ANOMALOUS DIFFUSION; ORDER; MEMORY; DISCRETIZATION; STABILITY; ACCURACY; SCHEMES;
D O I
10.1137/120880719
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We consider an initial-boundary value problem for partial derivative(t)u - partial derivative(-alpha)(t)del(2)u - f(t), that is, for a fractional diffusion (-1 < alpha < 0) or wave (0 < alpha < 1) equation. A numerical solution is found by applying a piecewise-linear, discontinuous Galerkin (DG) method in time combined with a piecewise-linear, conforming finite element method in space. The time mesh is graded appropriately near t = 0, but the spatial mesh is quasi-uniform. Previously, we proved that the error, measured in the spatial L-2-norm, is of order k(2+alpha-) + h(2)l(k), uniformly in t, where k is the maximum time step, h is the maximum diameter of the spatial finite elements, alpha(-) = min(alpha,0) <= 0, and l(k) = max(1, vertical bar log k vertical bar). Here, we prove convergence of order k(3+2 alpha-) l(k) + h(2) at each time level t(n) for -1 < alpha < 1. Thus, if -1/2 < alpha < 1, then the DG solution is superconvergent, which generalizes a known result for the classical heat equation (i.e., the case alpha = 0). A simple postprocessing step employing Lagrange interpolation leads to superconvergence for any t. Numerical experiments indicate that our theoretical error bound is pessimistic if alpha < 0. Ignoring logarithmic factors, we observe that the error in the DG solution at t = t(n), and after postprocessing at all t, is of order k(3+alpha-) + h(2) for -1 < alpha < 1.
引用
收藏
页码:491 / 515
页数:25
相关论文
共 45 条
[1]  
[Anonymous], 2010, NONLINEAR PHYS SCI
[2]  
Chen CM, 2012, MATH COMPUT, V81, P345, DOI 10.1090/S0025-5718-2011-02447-6
[3]   NUMERICAL SCHEMES WITH HIGH SPATIAL ACCURACY FOR A VARIABLE-ORDER ANOMALOUS SUBDIFFUSION EQUATION [J].
Chen, Chang-Ming ;
Liu, F. ;
Anh, V. ;
Turner, I. .
SIAM JOURNAL ON SCIENTIFIC COMPUTING, 2010, 32 (04) :1740-1760
[4]   Convolution quadrature time discretization of fractional diffusion-wave equations [J].
Cuesta, E ;
Lubich, C ;
Palencia, C .
MATHEMATICS OF COMPUTATION, 2006, 75 (254) :673-696
[5]   A numerical method for an integro-differential equation with memory in Banach spaces: Qualitative properties [J].
Cuesta, E ;
Palencia, C .
SIAM JOURNAL ON NUMERICAL ANALYSIS, 2003, 41 (04) :1232-1241
[6]   A fractional trapezoidal rule for integro-differential equations of fractional order in Banach spaces [J].
Cuesta, E ;
Palencia, C .
APPLIED NUMERICAL MATHEMATICS, 2003, 45 (2-3) :139-159
[7]  
Cui M., NUMER ALGOR IN PRESS
[8]   Compact alternating direction implicit method for two-dimensional time fractional diffusion equation [J].
Cui, Mingrong .
JOURNAL OF COMPUTATIONAL PHYSICS, 2012, 231 (06) :2621-2633
[9]   Compact finite difference method for the fractional diffusion equation [J].
Cui, Mingrong .
JOURNAL OF COMPUTATIONAL PHYSICS, 2009, 228 (20) :7792-7804
[10]   TIME DISCRETIZATION OF PARABOLIC PROBLEMS BY THE DISCONTINUOUS GALERKIN METHOD [J].
ERIKSSON, K ;
JOHNSON, C ;
THOMEE, V .
ESAIM-MATHEMATICAL MODELLING AND NUMERICAL ANALYSIS-MODELISATION MATHEMATIQUE ET ANALYSE NUMERIQUE, 1985, 19 (04) :611-643