*** 9629076 Mukherjee This Small Business Innovation Research Phase II project will develop a next-generation Computer Aided Design (CAD) software package for three-dimensional (3D) stress analysis and optimal shape design of solid structural components. Although use of CAD is quite common, the currently used finite element method (FEM) based approach has serious shortcomings. The primary obstacle is the need to discretize (mesh) the entire domain of the 3D body into (volume) finite elements. This task, which cannot be reliably and efficiently automated for bodies of complex shape, becomes particularly onerous during optimal shape design where the shape of the body changes during successive design iterations and, therefore, the body must be remeshed during each iteration. Also, volume discretization is not naturally compatible with either solid modeling or with shape design algorithms. This project uses an entirely new computational engine - the boundary contour method (BCM). This approach only requires meshing of the bounding surface of a body and numerical evaluation of regular line integrals for 3D problems. Therefore, automatic meshing, both initially and between design iterations, becomes straightforward and the need to evaluate only regular 1D integrals makes the process very efficient. Also, the BCM approach, being surface based, is naturally compatible with both solid modeling and shape design algorithms. The feasibility of the BCM approach, for stress and sensitivity analysis in 3D linear elasticity, has been demonstrated in the Phase I project. The primary research tasks in Phase II, in addition to further code development and testing, are (a) coupling of the code with optimization software in order to carry out shape optimization, (b) adaptive meshing - both initially and during each design iteration and (c) development of a parallel version of the code for operation on a network of desk-top computers. If the research is successful, the resulting package is expected to be a significant advance in CAD technology. It should help to blur the current artificial and costly distinction between analysis and design engineers, greatly reduce product time to market and product development cost, and should be in great demand throughout the industrial sector. ***

Agency
National Science Foundation (NSF)
Institute
Division of Industrial Innovation and Partnerships (IIP)
Type
Standard Grant (Standard)
Application #
9629076
Program Officer
G. Patrick Johnson
Project Start
Project End
Budget Start
1997-09-01
Budget End
1999-08-31
Support Year
Fiscal Year
1996
Total Cost
$300,000
Indirect Cost
Name
Dehan Engineering Numerics
Department
Type
DUNS #
City
Ithaca
State
NY
Country
United States
Zip Code
14850