A FULLY SPECTRAL METHOD FOR HYPERBOLIC EQUATIONS
β Scribed by P. SHUKLA; V. ESWARAN; S. MURTY BHALLAMUDI
- Publisher
- John Wiley and Sons
- Year
- 1996
- Tongue
- English
- Weight
- 607 KB
- Volume
- 39
- Category
- Article
- ISSN
- 0029-5981
No coin nor oath required. For personal study only.
β¦ Synopsis
This investigation presents a fully spectral method for solving coupled hyperbolic partial differential equations. The spectral method is based on the Galerkin+ollocation technique. Two different preconditioners, the Preissmann and upyind schemes, are evaluated for their performance in solving the discretized equations. It has been found, for the cases considered, that the upwind scheme is a viable preconditioner for the fully spectral discretization of hyperbolic PDEs. Its performance as a preconditioner is in every way superior to that of the Preissmann scheme. It is established that the relative accuracy of different numerical solutions is reliably indicated by the root-mean-square average of their residuals obtained by the discretization. It is also established that the scheme gives much better accuracy than the finite-difference Preissmann scheme, for the same amount of computational effort, for both linear and non-linear problems.
π SIMILAR VOLUMES
## Abstract Recently, it is found that telegraph equation is more suitable than ordinary diffusion equation in modelling reaction diffusion for such branches of sciences. In this article, we propose a numerical scheme to solve the oneβdimensional hyperbolic telegraph equation using collocation poin
accuracy of the spatial discretization (global errors in the range 10 Οͺ9 -10 Οͺ11 ) is maintained while allowing for large A spectral discretization of the equal width equation (EWE) is presented. The method is shown to be convergent and nonlinearly stepsizes. stable. Time-stepping is performed with
## Abstract A modified backward difference time discretization is presented for Galerkin approximations for nonlinear hyperbolic equation in two space variables. This procedure uses a local approximation of the coefficients based on patches of finite elements with these procedures, a multidimension