FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1984
1984 - 0066.PDF
100 10 = 1,000 • ?100 Space Shuttle -DC-3 Wright Flyer 10 1900 1920 1940 1960 Year 1980 2 000 Windtunnel time, and hence cost, has increased dramatically since the Second World War. Nasa argues that whole-airframe computational aero dynamics will reverse both trends, reduce risk and result in better ens The next goal will be the purchase of the first of the two supercomputers, a Cray-2. These machines, built by Cray Research of Minneapolis, are just coming on line, and are capable of sustained operation at 250 megaflops, or 250 million operations per second. Cray supercomputers are acknowledged to be among the fastest in the business. But others still in devel opment may leapfrog Cray technology, and Nasa is structuring the NAS programme so that the second super computer, to be purchased in 1987, will incorporate the state of the art at that time. At the moment, this machine is called simply the Second High Speed Processor(HSP-2). The initial phase of the work will see installation of the first element or the enormous memory that will be needed. For 1985 this will comprise a store capable of holding 64 million 64-bit words. Two years later this will have been expanded to a capacity of 256 million 64-bit words. When it is linked to the Cray-2 by the hyperchannel databus, a sustained computing power of 1 gigaflop — a billion operations per second —will be achieved. This will create the most powerful system of its type in the world. Even this vast computing power will be insufficient to fully solve the Navier-Stokes equations for a complete airframe. The N-S equa tions account for almost every kind of aerodynamic disturbance, but they are so complex that their solution for an aircraft would require more than three million times the computing power of a mid- Seventies Class VI machine. A compro mise is therefore inevitable. For NAS, this is the "Reynolds averaged N-S solution", which waters down the computing speed needed to no more than 30 times that of a Class VI machine. This will enable calcu lation to be made of the viscous flow over the entire aircraft. Viscous flow covers such problems as separation of flows, stall, buffet and flutter, prediction of aero dynamic forces, and performance close to aerodynamic limits, as well as a wide range of powerplant aerodynamics. The problems that are solveable by today's generation of computers would also, of course, be covered. In recent years, this has been extended from the study of airflow during cruise only (where viscous effects are not dominant), to the computation of non-linear effects such as wave drag and transonic pressure loads. The first benefits of the NAS programme will be the much quicker solu tion of localised aerodynamic problems, such as engine-airframe integration. These are currently solved in eight hours by a Class Vl-type research computer— too long for practical everyday applica tion. The NAS computer will deal with a particular design case in 15min, instantly bringing the facility into industry's field of view. Graves quotes the Airbus A310 wing as just one example of current achievements in computational aerodynamics. Less windtunnel testing time meant that extra time was available to address complex interface problems. The result was a more EXOCET FOR SEA POWER aeros I '•' ',••••:'•:•«•:" '/' ''•••7:v:/>'"' '^X^S^'^'-' "'v, • ,;/J zmmmi&i tV ¥•>.'.<•''r.'.,V-' •»wsv t-^ :\.^:\ ^fe&iieiffl )G0 EXOCET missiles participate in the defence of 27 countries throughout the world
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
