[Defense] Structure Simplification Of Unstructured Quadrilateral Meshes
Wednesday, November 2, 2022
1:30 pm - 3:00 pm
In
Partial
Fulfillment
of
the
Requirements
for
the
Degree
of
Doctor
of
Philosophy
Muhammad
Naeem
Akram
will
defend
his
proposal
Structure
Simplification
of
Unstructured
Quadrilateral
Meshes
Abstract
Structured quadrilateral (quad) meshes are preferred in many engineering and medical applications due to their desired numerical properties. Despite the extensive research, automatic generation of quad- and hex- meshes with optimal structure remains a challenge. In-lieu-of generating a quad-mesh directly from an arbitrary input, structure simplification can be performed on quad-meshes obtained from triangular meshes. Several techniques utilizing local and global operations for structure simplification have been proposed in this regard. Despite promising results, robustness, scalability, feature preservation and limitations of both local and global simplifications call for a novel and comprehensive framework that addresses the aforementioned limiting factors.
In this work, we introduce a few semi-global simplification operations and devise a new framework that utilizes separatrix based operations along with local and global operations for structure simplification of planar and surface quadrilateral meshes. We also employ an automated singularity movement technique to further simplify the mesh structure. In addition, we propose an angle-based mesh optimization algorithm to enhance the quality of simplified meshes. We provide a comprehensive comparison of the results obtained through our simplification framework with existing techniques on a variety of models and demonstrate that our framework is robust and can effectively reduce the mesh singularities to a large extent while preserving the features of quad meshes with complex structures. We also attempt to extend the framework to address the structure simplification of unstructured hexahedral (hex-) meshes that contain more complex scenarios due to the additional dimension. Even though the initial simplification results on some simple hex-meshes are promising, more works need to be done to extend the operations used in quad-meshes simplification for hex-meshes.
Wednesday,
November
2,
2022
1:30PM
-
3:00PM
CT
Online
via
Dr. Guoning Chen, dissertation advisor
Faculty, students and the general public are invited.
