The research objective of this Faculty Early Career Development (CAREER) Program project aims at developing an efficient method for experimentation that will save time and money for conducting experiments. Design of experiments is an engineering-statistical tool for improving the quality of products with minimal cost. The research plan consists of three parts: physical experiments, computer experiments, and computer-aided physical experiments. Physical experiments are the usual hardware experiments, whereas computer experiments refer to those experiments that are performed in computers using engineering models and finite element analysis. The methods currently employed for these two types of experimentation differ widely. To remedy this situation a unified approach will be used, which in turn should lead to new discoveries in each type of experimentation methods. In particular, a new Bayesian methodology will be used for the design and analysis of experiments. This project also introduces a novel experimentation strategy known as computer-aided physical experiments. Physical experiments are expensive and time-consuming to conduct whereas computer experiments are cheaper and faster, but their validity depends on model assumptions. The new strategy combines the advantages of the two types of experiments and therefore will have a significant impact on industrial experimentation. The education plan is aimed at disseminating and popularizing the new methods in industries to ensure its long-term success.

The new methodology will make an immediate impact in US economy by helping automotive companies; semi-conductor, nano materials, and other hi-tech product industries to cut down their product development cycle and to improve quality. Although the main focus of this research is on industrial applications, the methodology is general and will help in accelerating discoveries in many scientific disciplines that use experimental investigation. The education plan will help develop a new workforce that will be instrumental in improving product quality, lowering manufacturing costs, and reducing wastes, thereby benefiting the society as a whole in the long-term.

Project Start
Project End
Budget Start
2005-07-01
Budget End
2011-09-30
Support Year
Fiscal Year
2004
Total Cost
$400,000
Indirect Cost
Name
Georgia Tech Research Corporation
Department
Type
DUNS #
City
Atlanta
State
GA
Country
United States
Zip Code
30332