Skip to Main Content (Press Enter)

Logo UNINSUBRIA
  • ×
  • Home
  • Corsi
  • Insegnamenti
  • Professioni
  • Persone
  • Pubblicazioni
  • Strutture
  • Terza Missione
  • Attività
  • Competenze

UNI-FIND
Logo UNINSUBRIA

|

UNI-FIND

uninsubria.it
  • ×
  • Home
  • Corsi
  • Insegnamenti
  • Professioni
  • Persone
  • Pubblicazioni
  • Strutture
  • Terza Missione
  • Attività
  • Competenze
  1. Pubblicazioni

Spectral analysis and multigrid methods for finite volume approximations of space-fractional diffusion equations

Articolo
Data di Pubblicazione:
2018
Abstract:
We consider a boundary value problem in weak form of a steady-state Riesz spacefractional diffusion equation (FDE) of order 2 - α with 0 < α < 1. By using a finite volume approximation technique on uniform grids, we obtain a large linear system, whose coefficient matrix can be viewed as the sum of diagonal matrices times dense Toeplitz matrices. We study in detail the hidden nature of the resulting sequence of coefficient matrices, and we show that they fall in the class of generalized locally Toeplitz (GLT) sequences. The associated GLT symbol is obtained as the sum of products of functions, involving the Wiener generating functions of the Toeplitz components and the diffusion coefficients of the considered FDE. By exploiting a few analytical features of the GLT symbol, we obtain spectral information used for designing efficient preconditioners and multigrid methods. Several numerical experiments, both in the 1D and 2D cases, are reported and discussed, in order to show the optimality of the proposed algorithms.
Tipologia CRIS:
Articolo su Rivista
Keywords:
Finite volume approximation; Fractional diffusion equations; Generalized locally toeplitz matrix-sequences; Multigrid methods; Preconditioning; Singular value/eigenvalue distribution
Elenco autori:
Donatelli, M.; Mazza, M.; Serra-Capizzano, S.
Autori di Ateneo:
Analisi numerica
DONATELLI MARCO
SERRA CAPIZZANO STEFANO
Link alla scheda completa:
https://irinsubria.uninsubria.it/handle/11383/2085958
Pubblicato in:
SIAM JOURNAL ON SCIENTIFIC COMPUTING
Journal
  • Dati Generali

Dati Generali

URL

https://epubs.siam.org/doi/pdf/10.1137/17M115164X
  • Accessibilità
  • Utilizzo dei cookie

Realizzato con VIVO | Designed by Cineca | 26.5.1.0