Location

Salt Lake Community College

Start Date

10-5-2004 9:00 AM

Description

Certain applications of the finite element method require hexahedral meshes for the underlying discretization. A procedure, known as THexing, which is guaranteed to produce an all-hex mesh is to begin with a tetrahedral mesh and then subdivide each element into four hexahedra. This research presents a method for improving the THex approach, known as Diced THexing, or DTHexing. The DTHex approach is based on general coarsening tools. An initial triangle surface mesh is coarsened and smoothed iteratively until a coarse mesh of reasonable quality is obtained. The volume is then easily meshed using a tetrahedral scheme, then refined using ’h’ type modifications. The goal of this method is to 1) improve the quality of elements in the finite element mesh and 2) decrease the number of overall nodes. The DTHex approach has been successful at improving models on biological meshes without increasing node count. This research was conducted using the CUBIT software.

Share

COinS
 
May 10th, 9:00 AM

Improved Hexahedral Meshing on Biological Models

Salt Lake Community College

Certain applications of the finite element method require hexahedral meshes for the underlying discretization. A procedure, known as THexing, which is guaranteed to produce an all-hex mesh is to begin with a tetrahedral mesh and then subdivide each element into four hexahedra. This research presents a method for improving the THex approach, known as Diced THexing, or DTHexing. The DTHex approach is based on general coarsening tools. An initial triangle surface mesh is coarsened and smoothed iteratively until a coarse mesh of reasonable quality is obtained. The volume is then easily meshed using a tetrahedral scheme, then refined using ’h’ type modifications. The goal of this method is to 1) improve the quality of elements in the finite element mesh and 2) decrease the number of overall nodes. The DTHex approach has been successful at improving models on biological meshes without increasing node count. This research was conducted using the CUBIT software.