OLD DOMINION FACULTY MEMBER JOINS IN NSF PROJECT
An Old Dominion University faculty member will join in a new National Science Foundation project to aid computer simulation of chemical reactors and the spread of wilderness wildfires.
David E. Keyes, director of the university's new Center for Computational Sciences and until recently the chair of the Mathematics and Statistics Department at Old Dominion, will serve as lead investigator of Old Dominion part of the project, estimated at $800,000 over five years.
The project is led by Carnegie-Mellon University, Pittsburgh, and also includes participants from Rice University, Houston; Sandia National Laboratory, Albuquerque; the Boeing Commercial Airplane Company, Seattle; and Fendell Associates, Los Angeles.
The combined project is worth approximately $2.5 million in NSF-administered funds, with contributions-in-kind from Boeing and Sandia.
Boeing is an industry leader in computer simulation and has collaborated in earlier NSF-funded projects at Old Dominion. Sandia, part of the U.S. Department of Energy's research complex, will perform experiments that will guide and verify the university-based research.
The project will create and apply algorithms and software tools for on-line simulations that continuously assimilate sensor data from dynamic physical processes and generate optimal strategies for their control.
A number of critical industrial, scientific and societal problems -- in aerodynamics, energy, geophysics, infrastructure, manufacturing, medicine, chemical process and environmental applications -- stand to benefit from this research, according to the NSF.
As demonstration projects, the researchers will model and develop control strategies for two complex systems, one man-made and one natural: chemical vapor deposition reactors and the spread of wilderness wildfires.
Though the areas of application are diverse, the underlying mathematics is very similar, according to Keyes.
This NSF project combines synergistically with the recently announced U.S. Department of Energy "Scientific Discovery through Advanced Computing" project led by Old Dominion.
Both projects require the solution of large-scale mathematical models based on partial differential equations (PDEs). As recently as five years ago, a single such computational analysis would have been considered a major achievement.
However, with the advent of massively parallel computers, the solution of PDEs can now be considered almost "routine."This allows scientists to consider new classes of algorithms for optimizing systems (such as chemical reactors) governed by PDEs. Typically, this requires a new PDE analysis at every step of the optimization process.
The grant is part of $156 million in new NSF grants aimed at ensuring U.S. leadership in computer-based research. The winning projects will receive funds from NSF's Information Technology Research (ITR) priority area, which spurs Fundamental research and innovative uses of IT in science and engineering.
The CMU/ODU/Rice team responded to the "Dynamic Data Driven Simulations"(DDDS) part of the ITR initiative, a new IT research area that focuses on real-time assimilation of data, which can refine the results of a simulation as it is being computed.
"NSF is proud to be a leader with these bold ITR projects," said NSF director Rita Colwell in a presentation today to the President's Information Technology Advisory Committee meeting, where she announced the awards. "Through long-term, high-risk research, we expect a wide range of positive results that will benefit the nation as a whole. Our objective is to support the development of software and IT services that will help scientists and engineers make the kind of discoveries that will eventually be applied by industry."