PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 5, 20260 citations

Pointwise convergence of the local discontinuous Galerkin method on a Shishkin mesh for 2D reaction-diffusion problems

View Full Paper
YCYao ChengXWXuesong WangMSMartin Stynes

Key Points

  • To establish pointwise convergence for the local discontinuous Galerkin method applied to 2D reaction-diffusion problems with boundary layers.
  • Applied a layer-upwind flux on a Shishkin mesh for the LDG method
  • Utilized energy-norm supercloseness to analyze convergence
  • Combined discrete Green’s functions and local analysis with a cut-off function
  • Handled coarse, anisotropic boundary layer, and fine corner meshes with different techniques
  • Obtained nearly optimal convergence rates disregarding logarithmic factors
  • Established the best known pointwise error bounds for the LDG method with boundary layers
  • Included numerical results to confirm the sharpness of the convergence rates

Abstract

A local discontinuous Galerkin (LDG) method, based on a recently-developed “layer-upwind” flux, is applied on a Shishkin mesh to solve a reaction-diffusion problem posed on the 2D unit square. Typical solutions of this problem have an exponential boundary layer along each edge of the square. Energy-norm supercloseness, discrete Green’s functions, and local analysis with a cut-off function are combined to establish pointwise convergence for the LDG solution in the entire domain; in this derivation the coarse mesh, the highly anisotropic boundary layer mesh, and the fine corner mesh are each handled with a different technique to gain the sharpest local results. The convergence rates obtained are nearly optimal (provided that one disregards logarithmic factors and error terms that are multiplied by the small singular perturbation parameter) and are the best pointwise error bounds currently known for the LDG method when solving 2D reaction-diffusion problems with boundary layers. Numerical results are also included to test the sharpness of our results.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Cheng et al. (2025) studied this question.

synapsesocial.com/papers/6984348bf1d9ada3c1fb2d10https://doi.org/10.1051/m2an/2025065/pdf
Ask AI
Helpful
Bookmark
Share
View Full Paper