THE VIRGINIAN-PILOT
Copyright (c) 1995, Landmark Communications, Inc.
DATE: Tuesday, January 3, 1995 TAG: 9501030096
SECTION: LOCAL PAGE: B1 EDITION: FINAL
SOURCE: BY JAMES SCHULTZ, STAFF WRITER
LENGTH: Medium: 74 lines
Everybody has designed an airplane at one time or another. It's simple, really: a sheet of paper that's folded just right and creased just so.
Of course, paper planes tossed into a breeze carry only the imaginations of their makers. Grown-up versions, says Osama Kandil, chairman of Old Dominion University's Aerospace Engineering Department, carry considerably more challenge.
Kandil and the graduate students he oversees are involved in a multi-year project to help both NASA and the Air Force figure out structural stresses and strains on two new airplanes that are likely to be flying 21st century skies. Rather than laboring long hours in laboratories or wind tunnels, though, the ODU researchers have turned to one of the most powerful machines known.
``Supercomputers are a must for dealing with these problems,'' Kandil said. ``They (the problems) have so many unknowns, in the range of five (million) to 10 million variables. You can use a supercomputer to optimize both the aerodynamics and structural performance of aircraft.''
Kandil and students develop computer code for and then study what happens to wings and fuselages as they ``fly'' computerized versions of the airplanes across virtual models of what's popularly known as the sound barrier. Students and professors link to supercomputers at NASA Langley Research Center in Hampton and NASA Ames in California over fiber optic lines, using brute computation to solve complex aerodynamic equations.
Kandil says that using supercomputers saves time and money. Designers are able to put computerized prototypes through their paces before the first piece of metal is cut. The traditional means of developing new aircraft involved building models first and then laboriously testing them in wind tunnels.
``In a wind tunnel you need a team of at least 10 to 12 people, in addition to model makers, as you modify the design of a wing or fuselage,'' Kandil said. ``One person can run this computer code. You can cut down the wind tunnel need drastically.''
In particular, the Air Force is interesting in identifying the aerodynamic forces that will affect its on-the-drawing-boards fighter, the F-22. By the beginning of the next century, this supermaneuverable or ``superman'' plane should be able fly circles around most other aircraft in existence.
For its part, NASA is evaluating designs for a proposed supersonic airplane that would carry passengers and freight, halving current travel times to Europe and Asia.
Thus far, Old Dominion's links to the NASA supercomputers for both projects have come free. While the university has received a combined $225,000 grant for the work from NASA, the Air Force and the Virginia Space Grant Consortium, ODU President James Koch contends the real value is the supercomputer time.
Price it out, Koch says, at 6,000 hours of supercomputer time at a private-sector value of $2,500 per hour, and the in-kind value is in the millions.
``None of the other institutions in Virginia use as much of the in-kind services as we do,'' Koch asserts. ``We're fortunate enough to have people supplying it to use.''
The new year, however, may bring a change in Old Dominion fortunes. NASA is weighing plans to charge universities for supercomputer use. Should that occur, Chairman Kandil says, ODU's aerodynamic research could slow to a crawl.
``We don't have the money to pay,'' he said. ``We would have to go back to (computer) workstations.'' Without supercomputer power, the ODU research would take at least seven times longer to complete - if it could be done at all. ILLUSTRATION: Color photo by Christopher Reddick, Staff
ODU Professor Osama Kandil, center, and graduate students Steve
Massey and Margaret Menzies are using a supercomputer to study
airplanes.
by CNB