September 26, 2005
The Department of Energy (DoE) announced recently that the Institute for Computational Engineering and Sciences (ICES) at The University of Texas at Austin will be awarded a $1.2 million research grant on computational modeling of multi-scale phenomena. The grant is part of a new DoE research initiative aimed at resolving major open problems connected with treating physical phenomena and engineered systems that involve wide disparities in scale. For example, semiconductor manufacturing involves processes of devices only hundreds of atoms in dimension which occur in a billionth of a second, but the manufacturing process may involve components that are a thousand times larger and the process may be a billion times slower. Methods for simulating such enormous variations in scale do not exist and their development presents one of the most formidable scientific and computational problems to be encountered in the 21st century.
According to Oden, “We have been working for several years on a major unsolved problem in computer simulation: the selection of computational models of physical events so as to adaptively control modeling errors, the inherent error in representing real-life events by mathematical abstractions. Our preliminary results in this area led us to new theories and methods for comparing and updating models of events occurring at different time and space scales. Hence, we uncovered a new approach to multi-scale modeling, and this was the basis of our proposed work."
DoE announced a national competition for proposals aimed at resolving the multi-scale paradox after three national workshops on the subject. The ICES award is one of only a dozen awarded universities nation-wide and this only such award granted in Texas. The UT Austin project will be led by Professor J. Tinsley Oden, Associate Vice President for Research and Director of ICES, and will focus on developing theory methods of multiscale modeling with applications to imprint lighography processes for semiconductors. Other collaborators on this project are: Drs. C. Grant Willson, Leszek Demkowicz, Jon Bass, Serge Prudhomme, and other graduate students.
