Learning What Makes Planes Crash

The final exam was a heap of charred and twisted wreckage. That pile of mangled metal had once been a high-flying twin-engine Piper Aztec, the sort of upscale airplane that private pilots used to aspire to before the advent of faster, sexier aircraft like Jetstream turboprops and Learjets.

On March 5, 1985, the plane crashed under mysterious circumstances in a field near Foss, Okla., killing the pilot.

The National Transportation Safety Board sent a team of expert investigators. The wreckage was studied in place, then collected and hauled away, eventually to be dumped unceremoniously on the concrete floor of a hangar at Will Rogers World Airport in Oklahoma City.

After weeks of study and analysis, the NTSB determined the probable cause of the crash.

Now it was up to seven of us students at the Federal Aviation Administration’s Transportation Safety Institute to figure out the same thing. We had about an hour.

In all, 28 students, including this reporter, were enrolled in a recent 10-day course on crash investigation at the FAA’s Mike Monroney Aeronautical Center here. Most were graduate engineers--men and women who make their living on the technical side of the aviation business.

The NTSB uses a “party system” to investigate most air crashes, drawing upon the knowledge and investigative skills of interested parties, as well as the expertise of its own staff.

In the 2 1/2-year probe of the TWA 800 crash, for example, the parties included the FBI, FAA, TWA, the pilots’ union and a host of manufacturers and subcontractors who had built the Boeing 747.

This recent class included FAA officials, people from Boeing Co., Dassault Falcon Jet, Raytheon Systems and Textron Lycoming and representatives of Delta Air Lines and American Trans Air.

Because the military investigates its own crashes, there were personnel from the Army, Air Force and Air National Guard. Because other countries draw on U.S. aviation expertise, there were students from the Thai Department of Aviation and Transport Canada.

Most of these students expect to someday participate directly in the investigation of an air crash. They are among about 1,500 people trained at the institute each year.

The classroom work began with lectures on things like turbocharger failures on reciprocating engines and the crystalline structure of steel and aluminum alloys.

Then came instruction on how the type of damage on a propeller blade can indicate how the engine was running; that crescent-shaped marks on a metal bar can signal whether it had been weakened by fatigue before it snapped; that steel cables feather like “horsetails” when stretched to the breaking point; that the fracture point of hot aluminum looks like the frayed end of an old broom, and the fracture point of cool aluminum looks like the lip of a cream pitcher.

There were slides and videotapes of crash sites, many of them grisly.

“Our students are going to have to face things like that when they get out in the field,” said Gene Doub, a former NTSB investigator and one of the institute’s top instructors. “You have to deal with death out there. There’s no way to avoid it.”

Next it was cockpit instruments: Even those severely damaged in a crash can reveal a lot.

Sometimes the needle on a dial will freeze where it was at the moment of impact, providing clues to things like engine performance and airspeed. Even when the needle breaks, its “slap” against the face of the dial on impact can leave a ghostly imprint.

If the filament of a light bulb is stretched, that usually means the bulb was hot--and on--when the plane crashed. Cold filaments usually shatter.

Also, fires leave important clues about their origin and intensity. Drugs, fatigue, boredom, mental stress and poor health can affect the performance of flight crews. And unexpected weather can turn a pleasant journey into a nightmare.

With all that information in hand, it was time for the final exam.

The students were divided into teams and marched a couple of blocks to the “boneyard,” a fenced-off area containing four separate piles of wreckage. Our team got the Aztec.

My assignment was to establish the “four corners” of the aircraft, making sure that all its parts were there. Missing parts would indicate that the plane had started breaking up before hitting the ground.

The plane was indeed all there. But pretty soon, Doub showed up to say that the left wing actually had been found several hundred yards short of where the rest of the plane had come down.

That seemed to indicate that the left wing had come off in flight, and damage indicated that it had rolled back over the top of the plane, clipping the tail as it ripped away.

“Broomstrawing!” one team member shouted, pointing to the spot below the left engine mount where the left wing spar had failed. Sure enough, the stub of the spar looked like the business end of a broom. That meant that the spar had gotten extremely hot before it failed. Directly above it was the left engine. Had the engine caught fire while the plane was in the air, weakening the spar until it failed?

The blades of the left propeller were bent straight back, evidence that it wasn’t turning when it hit the ground. That showed that the left engine probably wasn’t running at the moment of impact, and the gauges in the cockpit seemed to indicate the same thing. A fire could have prompted the pilot to shut the engine down.

Then someone spotted conclusive evidence that the engine had caught fire: splatters of melted aluminum on the left landing gear that could have blown there only while the Piper was still in the air.

Doub, who had been nudging us in the right direction, concurred. “This is a good one. The broomstrawing, the metal splatters on the tire--they’re all classic signatures of an in-flight fire.”

Doub said further examination would have shown that a fitting on the fuel line to the engine had not been tightened sufficiently, probably during maintenance. The line apparently leaked fuel that was ignited by a spark or the heat of the engine.

“I wasn’t that sure we could do it,” student Mark Flora said later. “But once we started touching the metal, picking up parts, I began to see how it would all fit together.”