Neural Nets for Mesh Assessment
Abstract
We investigate the construction, training and application of a neural net for assessing element shape quality of practical unstructured grids arising in mesh generation, adaptive refinement and moving grid applications. Results of numerical experiments are included to validate the process and demonstrate performance of the neural net for both triangulations in 2D and tetrahedral tessellation in 3D.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2005
- Accession Number
- ADA437983
Entities
People
- Graham F. Carey
- Saeed Iqbal
Organizations
- University of Texas at Austin