Introducing topography in a quadrangular adaptive meshing algorithm

Aurélie Scuotto and Marius Rapenne and Paul Cupillard and Guillaume Caumon. ( 2023 )
in: 2023 {RING} meeting, pages 14, ASGA

Abstract

Seismic wave simulations using the spectral element method (SEM) benefit from adaptive meshes which minimize the computation cost. In particular, RAPENNE ET AL. (2022) recently proposed an adaptive quadrangular-only meshing algorithm for SEM simulations in 2D smooth anisotropic media. This algorithm relies on a quadtree division of the physical domain, followed by a Laplacian smoothing to further optimize the quality and performance of the mesh. However, its implementation only handles a flat top boundary with limited lateral variations of seismic wave speed in the shallow part of the domain. In this work, we address these two limitations by introducing a topography while preserving a satisfying element shape and maintaining the local adaptivity of the algorithm. First, an initial mesh is built such that the overall distance of the points at top of this mesh to the desired topography is minimized. Then, these points are moved to fit the topography while damping the displacement over all or part of the domain. The results show that the mesh is able to follow weak topography while maintaining its adaptivity.

Download / Links

BibTeX Reference

@inproceedings{scuotto_introducing_RM2023,
 abstract = {Seismic wave simulations using the spectral element method (SEM) benefit from adaptive meshes which minimize the computation cost. In particular, RAPENNE ET AL. (2022) recently proposed an adaptive quadrangular-only meshing algorithm for SEM simulations in 2D smooth anisotropic media. This algorithm relies on a quadtree division of the physical domain, followed by a Laplacian smoothing to further optimize the quality and performance of the mesh. However, its implementation only handles a flat top boundary with limited lateral variations of seismic wave speed in the shallow part of the domain. In this work, we address these two limitations by introducing a topography while preserving a satisfying element shape and maintaining the local adaptivity of the algorithm. First, an initial mesh is built such that the overall distance of the points at top of this mesh to the desired topography is minimized. Then, these points are moved to fit the topography while damping the displacement over all or part of the domain. The results show that the mesh is able to follow weak topography while maintaining its adaptivity.},
 author = {Scuotto, Aurélie and Rapenne, Marius and Cupillard, Paul and Caumon, Guillaume},
 booktitle = {2023 {RING} meeting},
 language = {en},
 pages = {14},
 publisher = {ASGA},
 title = {Introducing topography in a quadrangular adaptive meshing algorithm},
 year = {2023}
}