wigstheone
Dec 6, 01, 10:52 am
INVESTIGATORS may not know for several months or even longer what caused American Airlines Flight 587 to crash shortly after takeoff from Kennedy International Airport on Nov. 12. But they are sure about one thing: the right engine of the Airbus A300 was about due for major maintenance. The engines, manufactured by the General Electric Company, were scheduled for an overhaul every 10,000 hours they were in service. The right engine on the plane had already logged 9,788 hours since its last overhaul, the left engine 684 hours.
Like automobiles that are scheduled for an oil change every 3,000 miles, aircraft maintenance is largely determined by hours flown. But as many drivers will attest, car troubles often don't conform to a schedule. The same holds true for aircraft and other mechanical devices that rely on time-based maintenance.
So what if maintenance was only required when there was actually something wrong with the automobile or aircraft? Many cars already have such diagnostic devices that activate the "check engine" light when something goes wrong. Now researchers are working on the next-generation systems that will use radarlike signals to detect when a structure or mechanical part is about to fail, and even be able to alert users as to whether immediate action is needed to correct the problem.
"The idea is to move from maintenance based on time to one based on the condition of the materials," said Douglas E. Adams, an assistant professor of mechanical engineering at Purdue University, who will present his research on structural-health monitoring at a conference on smart technology in Scotland next week. "Why service something just because the owner's manual says to service it? We all know that car maintenance is based on how hard you drive the vehicle, not how long you drive it."
The technology uses a network of devices attached to or imbedded in, say, an airplane or a car. The devices send out high-frequency signals in a 360-degree pattern. Sensors pick up the return signals and send the data to a black box that continually compares information from the sensors to determine whether a part is wearing out and, more important, how long it will last.
"The diagnosis part — being able to tell something is wrong — is doable," Dr. Adams said. "The prognosis part, where we're actually going to have a machine tell an Air Force pilot they have to land because a problem is imminent, is the tougher challenge and the piece of the puzzle that we're still working on."
http://www.nytimes.com/2001/12/06/technology/circuits/06NEXT.html
Like automobiles that are scheduled for an oil change every 3,000 miles, aircraft maintenance is largely determined by hours flown. But as many drivers will attest, car troubles often don't conform to a schedule. The same holds true for aircraft and other mechanical devices that rely on time-based maintenance.
So what if maintenance was only required when there was actually something wrong with the automobile or aircraft? Many cars already have such diagnostic devices that activate the "check engine" light when something goes wrong. Now researchers are working on the next-generation systems that will use radarlike signals to detect when a structure or mechanical part is about to fail, and even be able to alert users as to whether immediate action is needed to correct the problem.
"The idea is to move from maintenance based on time to one based on the condition of the materials," said Douglas E. Adams, an assistant professor of mechanical engineering at Purdue University, who will present his research on structural-health monitoring at a conference on smart technology in Scotland next week. "Why service something just because the owner's manual says to service it? We all know that car maintenance is based on how hard you drive the vehicle, not how long you drive it."
The technology uses a network of devices attached to or imbedded in, say, an airplane or a car. The devices send out high-frequency signals in a 360-degree pattern. Sensors pick up the return signals and send the data to a black box that continually compares information from the sensors to determine whether a part is wearing out and, more important, how long it will last.
"The diagnosis part — being able to tell something is wrong — is doable," Dr. Adams said. "The prognosis part, where we're actually going to have a machine tell an Air Force pilot they have to land because a problem is imminent, is the tougher challenge and the piece of the puzzle that we're still working on."
http://www.nytimes.com/2001/12/06/technology/circuits/06NEXT.html