Ph.D. Proposal Presentation by
Monday, October 24, 2005
(Dr. Kok-Meng Lee, Chair)
" Effects of Adaptive Discretization on Numerical Computation using Meshless Method with Live-object Handling Applications"
Computation accuracy of Finite Element Method (FEM) depends on the quality of the mesh. Recently meshless methods (MLM) that inherit many advantages of FEM have emerged and yet, they need no explicit mesh structure to discretize geometry. This important feature makes MLM an attractive tool for solving engineering problems, such as large deformation contact and fracture mechanics, where adaptive meshes are often needed.
To fully exploit the advantages of MLM, it is necessary to understand the discretization methods used by MLM. In this research, a general framework for MLM is developed for solving the engineering problems encountered in the automation of handling of live-objects. Two particular examples of interest are the analysis of large deformation contact and the electromagnetic field computation. The objectives of this thesis research are to study the effect of different discretization methods and to develop a method for adaptively assigning nodes. In order to accomplish the objectives, this proposed thesis is organized into three tasks: The first task develops different ways of discretizing the problem domain for meshless methods, including an adaptive discretization method. The second task will validate the discretization methods developed in Task 1, and investigate their effects on the computation of magnetic field. In the third (last) task, the above method will be used to solve a mechanical contact problem.
While the MLM develop here has an immediate application in the poultry industry for solving the large deformation contact and for design of magnetic actuators, we expect that the proposed method will have a broad spectrum of engineering applications.