Comparison study of numerical methods for solving the Allen-Cahn equation

Darae Jeong, Seunggyu Lee, Dongsun Lee, Jaemin Shin, Junseok Kim

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

The goal of this paper is to present a brief review and a critical comparison of the performance of several numerical schemes for solving the Allen-Cahn equation representing a model for antiphase domain coarsening in a binary mixture. Explicit, fully implicit, Crank-Nicolson, and unconditionally gradient stable schemes are considered. In this paper, we show the solvability conditions of the numerical schemes and the decreasing property of total energy using eigenvalues of the Hessian matrix of the energy functional. We also present the pointwise boundedness of the numerical solution for the Allen-Cahn equation. To compare the accuracy and numerical efficiency of these methods, numerical experiments such as traveling wave and motion by mean curvature are performed. Numerical results show that Crank-Nicolson and nonlinearly stabilized splitting schemes are almost close to the analytic solution. However, if a large time step is used in the numerical test, we have only two results with linearly and nonlinearly stabilized splitting schemes in spite of having large gaps between analytic solution and numerical results. The other numerical schemes except for linearly and nonlinearly stabilized splitting schemes have unstable results when large time step is used.

Original languageEnglish
Pages (from-to)131-136
Number of pages6
JournalComputational Materials Science
Volume111
DOIs
Publication statusPublished - 2016 Jan 1

Fingerprint

Allen-Cahn Equation
Coarsening
Binary mixtures
Numerical methods
Numerical Methods
eccentrics
Numerical Scheme
Crank-Nicolson
Analytic Solution
Hessian matrices
antiphase boundaries
Linearly
Experiments
Motion by Mean Curvature
traveling waves
Numerical Results
Solvability Conditions
binary mixtures
Hessian matrix
Binary Mixtures

Keywords

  • Allen-Cahn equation
  • Crank-Nicolson
  • Explicit
  • Finite difference
  • Implicit
  • Unconditionally gradient stable

ASJC Scopus subject areas

  • Materials Science(all)
  • Chemistry(all)
  • Computer Science(all)
  • Physics and Astronomy(all)
  • Computational Mathematics
  • Mechanics of Materials

Cite this

Comparison study of numerical methods for solving the Allen-Cahn equation. / Jeong, Darae; Lee, Seunggyu; Lee, Dongsun; Shin, Jaemin; Kim, Junseok.

In: Computational Materials Science, Vol. 111, 01.01.2016, p. 131-136.

Research output: Contribution to journalArticle

Jeong, Darae ; Lee, Seunggyu ; Lee, Dongsun ; Shin, Jaemin ; Kim, Junseok. / Comparison study of numerical methods for solving the Allen-Cahn equation. In: Computational Materials Science. 2016 ; Vol. 111. pp. 131-136.
@article{515fef814081437487cccbb58a64ef04,
title = "Comparison study of numerical methods for solving the Allen-Cahn equation",
abstract = "The goal of this paper is to present a brief review and a critical comparison of the performance of several numerical schemes for solving the Allen-Cahn equation representing a model for antiphase domain coarsening in a binary mixture. Explicit, fully implicit, Crank-Nicolson, and unconditionally gradient stable schemes are considered. In this paper, we show the solvability conditions of the numerical schemes and the decreasing property of total energy using eigenvalues of the Hessian matrix of the energy functional. We also present the pointwise boundedness of the numerical solution for the Allen-Cahn equation. To compare the accuracy and numerical efficiency of these methods, numerical experiments such as traveling wave and motion by mean curvature are performed. Numerical results show that Crank-Nicolson and nonlinearly stabilized splitting schemes are almost close to the analytic solution. However, if a large time step is used in the numerical test, we have only two results with linearly and nonlinearly stabilized splitting schemes in spite of having large gaps between analytic solution and numerical results. The other numerical schemes except for linearly and nonlinearly stabilized splitting schemes have unstable results when large time step is used.",
keywords = "Allen-Cahn equation, Crank-Nicolson, Explicit, Finite difference, Implicit, Unconditionally gradient stable",
author = "Darae Jeong and Seunggyu Lee and Dongsun Lee and Jaemin Shin and Junseok Kim",
year = "2016",
month = "1",
day = "1",
doi = "10.1016/j.commatsci.2015.09.005",
language = "English",
volume = "111",
pages = "131--136",
journal = "Computational Materials Science",
issn = "0927-0256",
publisher = "Elsevier",

}

TY - JOUR

T1 - Comparison study of numerical methods for solving the Allen-Cahn equation

AU - Jeong, Darae

AU - Lee, Seunggyu

AU - Lee, Dongsun

AU - Shin, Jaemin

AU - Kim, Junseok

PY - 2016/1/1

Y1 - 2016/1/1

N2 - The goal of this paper is to present a brief review and a critical comparison of the performance of several numerical schemes for solving the Allen-Cahn equation representing a model for antiphase domain coarsening in a binary mixture. Explicit, fully implicit, Crank-Nicolson, and unconditionally gradient stable schemes are considered. In this paper, we show the solvability conditions of the numerical schemes and the decreasing property of total energy using eigenvalues of the Hessian matrix of the energy functional. We also present the pointwise boundedness of the numerical solution for the Allen-Cahn equation. To compare the accuracy and numerical efficiency of these methods, numerical experiments such as traveling wave and motion by mean curvature are performed. Numerical results show that Crank-Nicolson and nonlinearly stabilized splitting schemes are almost close to the analytic solution. However, if a large time step is used in the numerical test, we have only two results with linearly and nonlinearly stabilized splitting schemes in spite of having large gaps between analytic solution and numerical results. The other numerical schemes except for linearly and nonlinearly stabilized splitting schemes have unstable results when large time step is used.

AB - The goal of this paper is to present a brief review and a critical comparison of the performance of several numerical schemes for solving the Allen-Cahn equation representing a model for antiphase domain coarsening in a binary mixture. Explicit, fully implicit, Crank-Nicolson, and unconditionally gradient stable schemes are considered. In this paper, we show the solvability conditions of the numerical schemes and the decreasing property of total energy using eigenvalues of the Hessian matrix of the energy functional. We also present the pointwise boundedness of the numerical solution for the Allen-Cahn equation. To compare the accuracy and numerical efficiency of these methods, numerical experiments such as traveling wave and motion by mean curvature are performed. Numerical results show that Crank-Nicolson and nonlinearly stabilized splitting schemes are almost close to the analytic solution. However, if a large time step is used in the numerical test, we have only two results with linearly and nonlinearly stabilized splitting schemes in spite of having large gaps between analytic solution and numerical results. The other numerical schemes except for linearly and nonlinearly stabilized splitting schemes have unstable results when large time step is used.

KW - Allen-Cahn equation

KW - Crank-Nicolson

KW - Explicit

KW - Finite difference

KW - Implicit

KW - Unconditionally gradient stable

UR - http://www.scopus.com/inward/record.url?scp=84942543826&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84942543826&partnerID=8YFLogxK

U2 - 10.1016/j.commatsci.2015.09.005

DO - 10.1016/j.commatsci.2015.09.005

M3 - Article

VL - 111

SP - 131

EP - 136

JO - Computational Materials Science

JF - Computational Materials Science

SN - 0927-0256

ER -