Thesis
Image-based boundary element computation of three-dimensional potential problems
Washington State University
Master of Science (MS), Washington State University
2008
Handle:
https://hdl.handle.net/2376/102792
Abstract
Image-based boundary element computation is a computer-aided engineering method for performing simulations based on scanning images of physical objects. In this research, an image-based boundary element computational workflow is developed by tightly integrating the steps of image scanning, mesh regularization and the boundary element method. Mesh quality evaluation and mesh regularization strategies were developed to prepare the scanned images for boundary element computation. Two kinds of potential problems, namely the thermal potential and the bio-potential problems, were investigated to examine the feasibility of the integrated image-based boundary element computation. For thermal potential problems, scan images were collected on objects of large scale from laser scanning and small scale from the micro-CT scanning. Boundary element computation was performed to simulate the heat conduction on the scanned models. Numerical accuracy and computation speed were investigated by comparing the boundary element-based computational scheme with the finite element-based scheme. For bio-potential problems, laser scanning was used to scan geometry information of a human head and a brain from anatomically realistic models. Boundary element computation on bio-electrical potential was performed to inversely compute the cortical potential from simulated Electroencephalography (EEG) measurement on the human scalp. Truncated-Singular Value Decomposition (T-SVD) was implemented to tackle the solution difficulty caused by ill-conditioned matrices. Parallel computing and block matrix computing were performed to improve the computational speed and the efficiency in computational resource usage. Numerical case studies were conducted to demonstrate the efficiency and accuracy of the image-based boundary element method. Our results show that the image-based boundary element method can be an effective and promising approach for many science research and engineering applications.
Metrics
2 File views/ downloads
8 Record Views
Details
- Title
- Image-based boundary element computation of three-dimensional potential problems
- Creators
- Hui Zhang
- Contributors
- Xiaolin Chen (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; Pullman, Wash. :
- Identifiers
- 99900525105901842
- Language
- English
- Resource Type
- Thesis