NASA has completed a delicate set of flight tests to measure how modifications to an F-15 jet can affect the way shock waves form. The results could help turn sonic booms into distant rumbles.

The measurements will be used to calibrate a computer model of shock wave propagation which will be a crucial aid for engineers designing a new generation of quieter supersonic aircraft. "We're pretty close to being able to control sonic booms," says Peter Cohen of NASA's Langley Research Center in Virginia, principal investigator for the agency's supersonic research programme.

Shock waves form at the front and back of supersonic aircraft as they shove air out of the way. When these shock waves hit the ground, observers hear them as a single boom. Public opposition to booms has led to a ban on civilian supersonic flight over US land, and this key factor has discouraged further development of supersonic planes.

Attempts to quieten the sound have focused on a technique called boom shaping, which has it that booms would be weaker if they spread out over a larger area. The idea is to redesign the shock-forming zones around the nose and tail so that waves from them don't hit the ground together, but instead arrive over a longer period, like distant thunder.

Five years ago, NASA, the US Defense Advanced Research Projects Agency and several aerospace companies studied how a spiked nose-cone fitted to an F-5 fighter jet affected shock waves generated at the nose.

Now NASA has finished a similar set of measurements of the boom generated at the rear of an aircraft. This depends on the shape of the wings and air flow around the engine. The "Lift and Nozzle Change Effects Tail Shock" project (LANCETS) measured the shock waves produced by a modified F-15 jet in different wing and engine configurations. The measurements were taken by a second F-15 flying close behind and will be used to calibrate NASA's computer models of how shock waves form.

If the results live up to expectations, the next step would be to modify a jet so that it produces a low rumble rather than a boom in supersonic flight. Confirming the computer model is a crucial part of this process because designing, building and flying a test plane can cost more than $100 million, says Tim Moes of NASA's Dryden Research Center in California. "We think we could do it in the next four to five years," Cohen told New Scientist.

Don't expect to see a new Concorde any time soon, though. Cohen says that boom-hushing technology may first appear on small business jets rather than larger aircraft.

"Boom shaping doesn't come for free," says Cohen. The aerodynamic modifications reduce fuel efficiency by 10 to 15 per cent. "That's not too bad for a business jet, but not good for an airliner."