Implementation of an X-FEM solver for the classical two-phase Stefan problem (Q1930411): Difference between revisions
From MaRDI portal
Removed claim: author (P16): Item:Q849074 |
Changed an Item |
||
Property / author | |||
Property / author: Roland Griesse / rank | |||
Normal rank |
Revision as of 09:23, 21 February 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Implementation of an X-FEM solver for the classical two-phase Stefan problem |
scientific article |
Statements
Implementation of an X-FEM solver for the classical two-phase Stefan problem (English)
0 references
11 January 2013
0 references
The authors considered a two phase Stefan-problem in level-set formulation. They use discontinuous Galerkin (DG) scheme for the approximation of the level-set equation. The usage of quadratic ansatz spaces for both temperature and the level set equations lead to an accurate resolution of the curvature of the interface. Section 2 contains model equations and introduces the level-set equation in advective form. Extended finite element approximation with quadratic finite element functions and the use of the implicit Euler method in time are explained in the next section. The derivation of the fully discrete system is also included in this part. The weak form of the interface condition yields a linear system and is added as a penalty term into the fully discrete system. Section 4 contains an explanation of the extension of a recently introduced DG scheme for Hamilton-Jacobi equations on rectangular grids to triangular meshes in complex geometries. The DG spatial approximation is combined with strong stability preserving (SSP) Runge-Kutta methods in time. In the numerical examples of the final section, the authors use quadratic ansatz spaces and a third order SSP scheme. Section 5 deals with number of aspects of the implementation details including the location of the interface, numerical integration, extension velocity, evaluation of the Stefan condition, basis functions, narrow band level-set method, boundary conditions and the synchronization of the time-stepping schemes. In the final section, the authors investigated the order of convergence of their algorithm they introduced on the well-known Frank sphere example. For this example, the exact temperature and the interface location are known. Errors and the experimental order of convergence for the temperature and the interface positions are presented. The algorithm is also tested on a benchmark problem of a material solidifying in an infinite corner. The authors present examples of their method's capability of tracking of the interface accurately even in cases of complex geometries.
0 references
two-phase Stefan problem
0 references
solidification
0 references
moving boundary problem
0 references
level set method
0 references
discontinuous Galerkin approximation
0 references
extended finite element method
0 references