Back to Search Start Over

Numerical solution of time-dependent Emden-Fowler equations using bivariate spectral collocation method on overlapping grids

Authors :
Mkhatshwa Musawenkhosi P.
Motsa Sandile S.
Sibanda Precious
Source :
Nonlinear Engineering, Vol 9, Iss 1, Pp 299-318 (2020)
Publication Year :
2020
Publisher :
De Gruyter, 2020.

Abstract

In this work, we present a new modification to the bivariate spectral collocation method in solving Emden-Fowler equations. The novelty of the modified approach is the use of overlapping grids when applying the Chebyshev spectral collocation method. In the case of nonlinear partial differential equations, the quasilinearisation method is used to linearize the equation. The multi-domain technique is applied in both space and time intervals, which are both decomposed into overlapping subintervals. The spectral collocation method is then employed in the discretization of the iterative scheme to give a matrix system to be solved simultaneously across the overlapping subintervals. Several test examples are considered to demonstrate the general performance of the numerical technique in terms of efficiency and accuracy. The numerical solutions are matched against exact solutions to confirm the accuracy and convergence of the method. The error bound theorems and proofs have been considered to emphasize on the benefits of the method. The use of an overlapping grid gives a matrix system with less dense matrices that can be inverted in a computationally efficient manner. Thus, implementing the spectral collocation method on overlapping grids improves the computational time and accuracy. Furthermore, few grid points in each subinterval are required to achieve stable and accurate results. The approximate solutions are established to be in excellent agreement with the exact analytical solutions.

Details

Language :
English
ISSN :
21928010 and 21928029
Volume :
9
Issue :
1
Database :
Directory of Open Access Journals
Journal :
Nonlinear Engineering
Publication Type :
Academic Journal
Accession number :
edsdoj.f66dc2660fd946b5b49f988d24088feb
Document Type :
article
Full Text :
https://doi.org/10.1515/nleng-2020-0017