A study of the role of weights in the spectral properties of the least square finite difference scheme and the least square upwind finite difference scheme

被引:2
作者
Mungur, Maheshsingh [1 ]
Dauhoo, Muhammad Zaid [1 ]
机构
[1] Univ Mauritius, Dept Math, Reduit, Mauritius
关键词
finite difference; least squares; meshless scheme; power spectrum; spectral analysis; weight; POINT METHOD; MULTIGRID SOLUTION; EULER EQUATIONS; DISSIPATION; DISPERSION; CONVECTION; FLOW; ERROR;
D O I
10.1002/fld.5075
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
In this article, we analyze the performance of a family of nonlinear, shock-capturing mesh-free finite-difference schemes in wavenumber space. These finite difference methods are based on the use of a weighted least-square approximation procedure together with a Taylor series expansion of the unknown function. The influence of weights on the spectral properties of these schemes are studied using the approximate dispersion relation. The numerical study evidences that an appropriate choice of the weighting functions reduces dissipation and dispersion inherent to the schemes. The meshless schemes studied in this work, are then used to simulate 2D internal inviscid subsonic and transonic flows respectively in a channel with a bump. The results in 2D corroborate with the 1D analyses and reveal that when the supporting nodes closer to the reference node are assigned a greater weightage value in the computation of the 1st order spatial derivatives, better spectral properties are achieved at low wavenumber. However, the present numerical study shows that weighting function does not shift the frequency of the errors and thus in order to accelerate the convergence of these schemes, methods such as the classic multigrid can be used.
引用
收藏
页码:803 / 848
页数:46
相关论文
共 59 条
[51]   Spectral analysis of finite difference schemes for convection diffusion equation [J].
Suman, V. K. ;
Sengupta, Tapan K. ;
Prasad, C. Jyothi Durga ;
Mohan, K. Surya ;
Sanwalia, Deepanshu .
COMPUTERS & FLUIDS, 2017, 150 :95-114
[52]   A critical analysis of some popular methods for the discretisation of the gradient operator in finite volume methods [J].
Syrakos, Alexandros ;
Varchanis, Stylianos ;
Dimakopoulos, Yannis ;
Goulas, Apostolos ;
Tsamopoulos, John .
PHYSICS OF FLUIDS, 2017, 29 (12)
[53]  
Thomas J.W., 2013, Numerical partial differential equations: finite difference methods, V22
[54]   Optimal Runge-Kutta schemes for discontinuous Galerkin space discretizations applied to wave propagation problems [J].
Toulorge, T. ;
Desmet, W. .
JOURNAL OF COMPUTATIONAL PHYSICS, 2012, 231 (04) :2067-2091
[55]   GROUP-VELOCITY IN FINITE-DIFFERENCE SCHEMES [J].
TREFETHEN, LN .
SIAM REVIEW, 1982, 24 (02) :113-136
[56]   Optimized Runge-Kutta methods with minimal dispersion and dissipation for problems arising from computational acoustics [J].
Tselios, Kostas ;
Simos, T. E. .
PHYSICS LETTERS A, 2007, 363 (1-2) :38-47
[57]  
Vichnevetsky R., 1982, FOURIER ANAL NUMERIC
[58]   LSFD method for accurate vibration modes and modal stress-resultants of freely vibrating plates that model VLFS [J].
Wang, C. M. ;
Wu, W. X. ;
Shu, C. ;
Utsunomiya, T. .
COMPUTERS & STRUCTURES, 2006, 84 (31-32) :2329-2339
[59]   Weighted Least-Squares Finite Element Method for Cardiac Blood Flow Simulation with Echocardiographic Data [J].
Wei, Fei ;
Westerdale, John ;
McMahon, Eileen M. ;
Belohlavek, Marek ;
Heys, Jeffrey J. .
COMPUTATIONAL AND MATHEMATICAL METHODS IN MEDICINE, 2012, 2012