Application of the Chebyshev collocation method to solve boundary value problems of heat conduction
Journal: Discrete and Continuous Models and Applied Computational Science (Vol.32, No. 1)Publication Date: 2024-04-01
Authors : Konstantin Lovetskiy; Stepan Sergeev; Dmitry Kulyabov; Leonid Sevastianov;
Page : 74-85
Keywords : initial boundary problems; pseudo spectral collocation method; Chebyshev polynomials; Gauss-Lobatto sets; numerical stability; separation of variables;
Abstract
For one-dimensional inhomogeneous (with respect to the spatial variable) linear parabolic equations, a combined approach is used, dividing the original problem into two subproblems. The first of them is an inhomogeneous one-dimensional Poisson problem with Dirichlet-Robin boundary conditions, the search for a solution of which is based on the Chebyshev collocation method. The method was developed based on previously published algorithms for solving ordinary differential equations, in which the solution is sought in the form of an expansion in Chebyshev polynomials of the 1st kind on Gauss-Lobatto grids, which allows the use of discrete orthogonality of polynomials. This approach turns out to be very economical and stable compared to traditional methods, which often lead to the solution of poorly defined systems of linear algebraic equations. In the described approach, the successful use of integration matrices allows complete elimination of the need to deal with ill-conditioned matrices. The second, homogeneous problem of thermal conductivity is solved by the method of separation of variables. In this case, finding the expansion coefficients of the desired solution in the complete set of solutions to the corresponding Sturm-Liouville problem is reduced to calculating integrals of known functions. A simple technique for constructing Chebyshev interpolants of integrands allows to calculate the integrals by summing interpolation coefficients.
Other Latest Articles
- Solving the eikonal equation by the FSM method in Julia language
- Sampling of integrand for integration using shallow neural network
- Chronology of the development of Active Queue Management algorithms of RED family. Part 2: from 2006 up to 2015
- Statistical methods for estimating quartiles of scientific conferences
- Low Costs Electrical Calibration System of SLM with the Uncertainty Measurements Compared with Primary System Platform Brūel & Kjær type 3630
Last modified: 2024-07-19 21:41:46