Extending meshless method of approximate particular solutions (MAPS) to two-dimensional convection heat transfer problems
•Numerical schemes are developed to widen the application of MAPS.•MAPS is implemented to solve mixed heat convection problems.•Fluid flow and heat transfer are studied using two kinds of particular solutions. Natural, forced and mixed convection heat transfer problems are solved by the meshless Met...
Gespeichert in:
Veröffentlicht in: | Applied mathematics and computation 2021-02, Vol.390, p.125484, Article 125484 |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | •Numerical schemes are developed to widen the application of MAPS.•MAPS is implemented to solve mixed heat convection problems.•Fluid flow and heat transfer are studied using two kinds of particular solutions.
Natural, forced and mixed convection heat transfer problems are solved by the meshless Method of Approximate Particular Solutions (MAPS). Particular solutions of Poisson and Stokes equations are employed to approximate temperature and velocity, respectively. The latter is used to obtained a closed expression for pressure particular solution. In both cases, the source terms are multiquadric radial basis functions which allow to obtain analytical expressions for these auxiliary problems. In order to couple momentum and energy equations, a relaxation strategy is implemented to avoid convergence problems due to the difference between successive temperature and velocity changes when solving the steady problem from an initial guess. The developed and validated numerical scheme is used to study flow and heat transfer in two two-dimensional problems: natural convection in concentric annulus between a square and a circular cylinder and non-isothermal flow past a staggered tube bundle. Numerical solutions obtained by MAPS are comparable in accuracy to solutions reported by authors who uses denser nodal distributions, showing the capability of the present method to accurately solve heat convection problems with temperature and heat flux boundary conditions as well as curve geometries and internal flow situations. |
---|---|
ISSN: | 0096-3003 1873-5649 |
DOI: | 10.1016/j.amc.2020.125484 |