A Review and Implementation of the Generalized Multiscale Method (GMsFEM)

Authors

Keywords:

partial differential equations, finite element method, multiscale method, generalized multiscale method, diffusion problems, heterogeneous media

Abstract

The aim of this article is to study and apply three numerical methods for solving elliptic partial differential equations: the finite element method (FEM), the multiscale finite element method (MsFEM), and the generalized multiscale finite element method (GMsFEM), with special emphasis on the latter. To this end, a theoretical review of their variational and matrix foundations is presented, followed by computational implementations for homogeneous, heterogeneous, and high-contrast diffusion problems. The simulations were carried out using FreeFem++ and MATLAB, which made it possible to compare the performance of the methods under different scenarios. The results show that the classical multiscale method has limitations when the problem involves highly heterogeneous media or non-separable scales, whereas the generalized multiscale method provides a better representation of local behavior through the construction of basis functions obtained from local spectral problems. It is concluded that this latter method is a more suitable alternative for the numerical approximation of complex multiscale problems, especially in high-contrast settings, where reducing degrees of freedom is required without significantly compromising approximation quality.

Abstract
1
PDF (Spanish)
0

Author Biographies

A. Forero-Laiton, National University of Colombia – Bogotá Campus. Bogotá, Colombia.

Department of Mathematics, National University of Colombia – Bogotá Campus. Bogotá 111321, Colombia.

Y. Sarmiento-Perdomo, National Autonomous University of Honduras – University City. Tegucigalpa, Honduras.

School of Mathematics and Computer Science, Faculty of Sciences, National Autonomous University of Honduras – University City. Tegucigalpa 11101, Honduras.

Y. Trujillo-Ladino, National University of Colombia – Bogotá Campus. Bogotá, Colombia.

Department of Mathematics, National University of Colombia – Bogotá Campus. Bogotá 111321, Colombia.

Downloads

Published

2026-06-24

How to Cite

A Review and Implementation of the Generalized Multiscale Method (GMsFEM). (2026). Portal De La Ciencia, 1(21), 11-38. https://doi.org/10.5377/pc.v1i21.23092

Issue

Section

Original scientific article

How to Cite

A Review and Implementation of the Generalized Multiscale Method (GMsFEM). (2026). Portal De La Ciencia, 1(21), 11-38. https://doi.org/10.5377/pc.v1i21.23092