The center will advance the emerging field of "predictive science," or applying computational simulations to predict the behavior of complex systems, said Jayathi Y. Murthy, director of the new center and a professor in Purdue's School of Mechanical Engineering.
PRISM and the other four newly selected centers will focus on unclassified applications of interest to NNSA and its three national laboratories: Lawrence Livermore, Los Alamos and Sandia.
Under PRISM, the miniature switches, called MEMS devices, are being created to replace conventional switches and other electronic components. MEMS are machines that combine electronic and mechanical components on a microscopic scale.
The MEMS are far lighter and smaller than the conventional technology and could be manufactured in large quantities at low cost, Murthy said.
"Research is needed, however, to improve the reliability, ruggedness and durability of the devices," she said.
The new simulations will make it possible to accurately predict how well the MEMS devices would stand up to the rigors of varying and extreme environments and how long they would last in the field. Devices in many environments must withstand crushing gravitational forces, temperature extremes, radiation and shocks from impact.
"Reliability pertains to long-term performance," Murthy said. "Improving the integrity and survivability relate to the fact that MEMS get used in very adverse conditions. You don't want the MEMS to fail before the systems in which they are embedded are deployed. MEMS have many potential important applications in civilian and defense applications."
For example, the switches can be used to turn radio signals on and off for a variety of purposes in national defense and for routing satellite communications. Potential civilian applications include cell phones and other telecommunications products, automotive sensors, and liquid-crystal-display projectors for large screens.
The technology will make it possible to reduce the size of switching equipment from several inches to 1 millimeter, or thousandth of a meter.
"Even though MEMS have a big size, weight and cost advantage, they are not really reliable enough yet," Murthy said.
A major challenge is creating "multiscale" simulations that bridge a broad range of size and time scales associated with objects measured in nanometers, or billionths of a meter, to objects measured in millimeters.
One problem is that matter behaves differently on the scale of nanometers than it does in the ordinary macro world of meters. Another complication is that important failure phenomena in MEMS may occur over a range of time scales, ranging from billionths of a second to several months.
The center will focus on creating simulations to unite these sizes and time scales, capturing the entire workings of a design, from its nanometer-scale layout to its macro-scale features. The research will draw on expertise and facilities affiliated with Purdue's Network for Computational Nanotechnology, based at the Birck Nanotechnology Center, and the Rosen Center for Advanced Computing, a division of Purdue's Office of Information Technology. The NNSA's national laboratory personnel will be advisers and collaborators in this research effort.
The research will concentrate on specific types of MEMS, called radio frequency MEMS, and particularly a device called a metal-dielectric contacting MEMS. The tiny switches have a length of about 400 microns, or millionths of a meter, or roughly four times the width of a human hair. The devices, switches used to turn on and off radio frequency signals, are made of a thin metal membrane located on top of a dielectric contact.
During operation, the membrane snaps on top of the contact, altering an electronic property called capacitance and switching off the radio signal, in effect turning off the device.
Researchers in the center will create a simulation system called MEMOSA to accurately model the devices. The metal membrane constantly hitting the contact forms cracks and defects. Whereas the defects are formed in regions a few hundred nanometers long, components in the device are 100 times larger, complicating the job of creating accurate simulations.
In addition to multiscale considerations, another complicating factor is that device operation involves the interaction of mechanical, electrical and thermal factors. The devices are made of various types of materials, which also have to be incorporated into simulations.
Creating the simulations will require the expertise of researchers from materials science, electrical engineering, mechanical engineering, aeronautics and astronautics, mathematics, computer science, and computer architecture.
The researchers will have access to unclassified supercomputers at the three NNSA national labs to run the large-scale simulations. These systems will be at the petascale computing level.
"Petascale computing is the leading edge, the fastest computing that will be possible in the near future," Murthy said. "Right now, the state of the art is terascale computing, a thousand times slower."
Researchers also will use computer resources on the nanoHUB, an Internet-based science gateway that provides access to advanced simulation and software tools. The nanoHub is part of the Network for Computational Nanotechnology at Purdue. Facilities and hardware provided by Purdue's Office of Information Technology also will be utilized extensively. ###
Writers: Emil Venere, (765) 494-4709, venere@purdue.edu, Phillip Fiorini, (765) 496-3133, (765) 427-3009 (cell), pfiorini@purdue.edu
Sources: France A. Cordóva, (765) 494-9708, Jayathi Y. Murthy, (765) 494-5701, jmurthy@ecn.purdue.edu, Jay Gore, interim director of Purdue's Energy Center, (765) 494-2122, gore@purdue.edu, Ananth Grama, Purdue professor of computer science, (765) 494-6964, ayg@purdue.edu
Related Web sites:
- Jayathi Y. Murthy
- Jayathi Y. Murthy
- Discovery Park
- Birck Nanotechnology Center
- Energy Center
- Network for Computational Nanotechnology
No comments:
Post a Comment