The Mystery Of Usair 427

Accidents: For the fifth time in five years, a USAir jet crashes -- this time in a nose dive near Pittsburgh that killed all 132 on board. There were questions about a troubled airline but few early clues to what went wrong.

Those who struggled into the wooded Pennsylvania ravine where USAir 427 disappeared last week were clearly shaken by what they saw. There was no plane -- no fuselage, no wings, no particularly identifiable chunk of what had once been a Boeing 737. There were no survivors, of course, and there were no bodies -- only body parts, bits and pieces of 132 human beings scattered through the shattered trees and all around the flaming crater where the plane had fallen. The firemen came and put out the fire. The medics came and gaped. "You want to have a feeling you can do something when you get there," said paramedic Steve Bailey. "Once you realize there's nothing you can do. . . . " His voice trailed off. "It blew people out of the plane -- blew them apart, into smithereens," said Beaver County Sheriff Frank Policaro Jr. "I have never seen mutilated bodies like that."

Apart from its grisly finality, two things distinguished USAir 427 from past commercial air disasters. Although the incidence of major air disasters has dropped somewhat since 1989, last week's tragedy was the fifth fatal crash in five years -- and the second in three months -- for financially troubled USAir. And the Pittsburgh crash was already shaping up, through the frenzied first days of a National Transportation Safety Board investigation (page 26), as a total mystery. Weather was not a factor, and there was no evidence of a bomb explosion or a close encounter with another plane. Pilot error, to judge by the radio traffic between the plane and the tower at Greater Pittsburgh International, did not seem to be a question. The 737 was only seven years old and had recently been serviced. All that was known, last week, was that Flight 427 made an unexplained roll, or yaw, to the left and almost instantly went out of control. "It wasn't a diving crash; it was a nose dive straight down -- a gigantic ball of fire," said Richard Trenary of Hopewell Township, Pa. "We heard a real loud, heavy thud. Then once it hit the ground, it exploded immediately."

It also rocked USAir, the Arlington, Va.-based company that is the sixth biggest U.S. domestic carrier. With more than 400 jets and gross revenues of $7.1 billion in 1993, USAir has been struggling in recent years to hold its own in the hotly competitive U.S. air-travel market. Though it reported a second-quarter profit this year, the company has lost $2.4 billion since 1989 and is struggling to maintain its cash flow. News of last week's disaster may hurt passenger bookings, especially since the company suffered another major crash, with 37 fatalities, in Charlotte, N.C., on July 2. "It's scary," said Arnold Barnett, an MIT air-safety expert and research consultant for Newsweek. "I really don't believe that some U.S. carriers are better than others. But at the moment, to be honest, I'm just heartsick about USAir and I don't know what to make of it. I can't bring myself to say, "This is just bad luck'."

In an interview with Newsweek, USAir CEO Seth Schofield pleaded with the public "not to paint the airline into a corner it doesn't deserve." Schofield acknowledged USAir's urgent need to cut costs but insisted "unequivocally" that the company could do so without compromising safety. "I think the public should hear that independent sources are validating what we are saying -- that USAir is a safe airline," he said. "Each of [the five] accidents has been very dissimilar." That is essentially true. Although federal safety officials have not announced their findings on the July crash, it appears to have been weather-related. Another crash, at Los Angeles International Airport in 1991, is attributed to a mistake by an FAA air-traffic controller. A 1989 crash at New York's La Guardia Airport was found to be pilot error, and another crash at La Guardia, in 1992, was attributed to lax industry standards on de-icing a plane's wings in winter weather.

None of this seems germane to last week's crash in Pennsylvania. Flight 427 lifted off from Chicago's O'Hare airport just before 5 p.m. Central time last Thursday. The 737-300 -- known to pilots as "Fat Albert" for its plump fuselage and stubby length -- was filled to capacity for the 429-mile flight to Pittsburgh, a major hub for USAir. Later, NTSB investigators would attempt to trace down a report that a passenger on the same plane reported a loud engine noise earlier in the day -- but if there was any relevance to the rumor, no one knew what it was. The Chicago-Pittsburgh flight was routine, with ideal weather conditions, until the very end.

Shortly before 7 p.m., pilot Peter Germano, 45, brought the plane down to 5,800 feet to begin his final approach to Pittsburgh International, about six miles away. A transcript of the plane's radio transmissions reveals routine communication between Pittsburgh tower and USAir 427 -- until the moment that Germano and his copilot, Charles B. Emmett III, 38, realized something was terribly wrong. There is still no way to pinpoint the problem. Although the tape shows that Germano and Emmett were alerted to the presence of another plane -- a twin-prop Grumman Jetstream -- NTSB officials and Pittsburgh air-traffic controllers say the Jetstream was miles away from the 737 and probably did not pose a risk.

USAir 427, it will be [runway] two-eight right.

Two-eight right. Thank you.

USAir 427 turn left, heading one-zero-zero. Traffic will be [at] one to two o'clock -- six miles northbound, Jetstream climbing out of 33 [3,300] for 5,000.

We're looking for the traffic. Turning to one-zero-zero. USAir 427.

Oh! [Unintelligible] Oh, God!

USAir 427 maintain 6,000, over.

[Unintelligible] Traffic emergency. [Unintelligible] Oh, s--t! Ahhh. [Unintelligible] Ahhhh! Ahhh! Ahhhh!

This death-dive lasted roughly 23 seconds -- undoubtedly moments of stark, screaming terror for everyone aboard, but one that ended with merciful suddenness when the plane hit the ground. Many air disasters -- certainly those that occur during the final descent toward the airport -- lead to a skimming, glancing impact that leaves the fuselage mostly intact. That is why at least some passengers survive. But USAir 427 dove very steeply and very fast -- more than 275 miles per hour, according to the NTSB. The plane at that point "was a bomb," one aviation expert said. It's basic physics, he said -- "velocity squared times weight, and all that fuel on board." He theorized that no one aboard felt anything as the plane hit the ground because of the nearly instantaneous "vaporization time." "You're living," he said, "and then -- poof! -- you're not."

Beaver County coroner Wayne Tatalovich said it was probably impossible to identify the remains of everyone on board -- and it may be just as difficult for the NTSB to determine the cause of the crash. Scores of people saw Flight 427's horrifying last moments, and at the weekend, NTSB teams were beginning to interview at least some of them. Investigators have learned, however, that eyewitness accounts can be misleading, and this crash seemed to fit the historical pattern. For example, news accounts quoted some who thought the plane had had an engine problem: Richard Trenary heard a sound "like a muffled backfire" and saw "little puffs of smoke" coming from the left engine. Others reported a roll to the right, and a 10-year-old named Jason Moka said, "There was fire and smoke coming out of [the plane]. Then it turned upside down and spun."

NTSB officials seemed to be discounting the talk of fire and smoke, and they said the plane's flight data recorder showed a roll to the left -- not to the right. More importantly, speculation by pilots and aviation experts interviewed by Newsweek seemed to focus on the possibility of a catastrophic failure of some element in the plane's control systems. The reason was Flight 427's wild, steep dive from 6,000 feet to the ground. "I don't think engine failure could have caused this," said retired USAir Capt. Ted Maselko. "Engine failure is almost [routine] to pilots -- it's what they're trained for. If one engine cuts out, you just put more power on the other." Hubert Smith, an aerospace-engineering professor at Pennsylvania State University, agreed. "Even if the airplane lost both engines, it would still fly," he said.

Another USAir pilot, Capt. Morey Stein, compared Flight 427 to the crash of a United Airlines 737 at Colorado Springs in 1991. "The 737 is a very good, very docile airplane," Stein said. Aside from sabotage or terrorism, he said, Flight 427 must have suffered "some control malfunction that happens only extremely rarely." The Colorado Springs crash also is a mystery. Last year the NTSB issued a report narrowing its probable causes to two: a freak weather effect known as a "rotor," which is something like a horizontal tornado, or control failure that is known, in flying jargon, as a "hard-over runaway rudder."

Investigators will clearly be looking for similarities between the Pittsburgh and Colorado Springs crashes. One potential culprit is what is known as the power control unit, a hydraulic system that governs the plane's rudder and elevators. The rudder is the vertical flap on the tail; it makes the airplane turn. The elevators are the horizontal flaps on a plane's rear wings; they make the aircraft rise or dive. It is unlikely that Flight 427 would have gone into so steep and sustained a dive without some malfunction of the elevator controls. Though there are clear differences between the two crashes, both 737s suffered an unexplained roll just moments before they began to dive. Another suspect is what are known as "thrust reversers" on the plane's jet engines. Thrust reversers are used to slow the plane after it has touched down. But NTSB officials said at least one of Flight 427's 12 thrust reversers had been deployed; if others had been deployed as well, they said, that could have had "catastrophic consequences" to the plane's stability.

It may be many months before NTSB investigators find the cause of Flight 427's demise -- if indeed they ever do. "You can't rush to judgment on what caused this thing," said Aaron Gellman, director of Northwestern University's Transportation Center. "It's extremely complicated -- there are so many interconnecting technologies and interacting people. You've got to reconstruct everything that happened during this flight." One reason for caution is that an NTSB report implicating the 737's control systems could lead to multimillion-dollar damage suits against the Boeing Co. or its subcontractors. "At this point, it's all speculation," said Boeing spokesman Jerry Johnson. "It's too early to know what may have caused the incident." Asked about possible rudder problems with the 737, Johnson made the same point again: "It's premature to discuss any of those issues now."

Still, the odds are improving that the questions about Flight 427 will eventually be answered. That's because the flight data recorder on the USAir 737 was a new-model device that collects far more information on the plane's control systems than the "black box" on the United plane that crashed at Colorado Springs. This new recorder should enable the NTSB to re-create the plane's final moments with considerable precision -- and possibly, solve the mystery of its terrifying dive.

Countries whose airlines do Countries with not meet FAA standards conditional acceptance Belize Bolivia Dominican Republic El Salvador Gambia Guatemala Ghana Netherlands Antilles Honduras Nicaragua Paraguay Uruguay Zaire Source: The Washington Post

Death risk Other major per jet flight USAir U.S. carriers 1977-86 0 1 in 9 million 1987-90 1 in 3 million 1 in 10 million 1991-94 1 in 1 million 1 in 15 million Sources: Arnold Barnett, MIT. From FAA DATA

Fatal Accidents Fatalities 1981 4 4 1982 3 233 1983 4 15 1984 1 4 1985 4 197 1986 1 3 1987 4 231 1988 3 285 1989 8 131 1990 6 39 1991 4 62 1992 4 33 1993 0 0 1994 2 169 *U.S. scheduled commercial airline flights, does not include commuter airlines. Sources: Statistical abstract, accident facts

United States airlines Death risk per flight* USAir 1 in 2.5 million Northwest 1 in 4 million TWA 1 in 5 million United 1 in 8 million American 1 in 10 million Delta 1 in 10 million Continental 1 in 15 million Southwest 0 (perfect record) Entire group 1 in 7 million International airlines Death risk per flight* Aeroflot** 1 in 200,000 Air-India 1 in 200,000 EgyptAir 1 in 200,000 Malev (Hungary) 1 in 200,000 Varig 1 in 200,000 Aeromexico 1 in 300,000 Air Canada 1 in 1.5 million Japan Airlines 1 in 2 million British Airways 1 in 3 million Lufthansa 1 in 5 million El Al 1 in 20 million Argentinas 0 Singapore 0 Industrialized countries 1 in 3.5 million (21 nations) Developing countries 1 in 500,000 (61 nations) *This statistic is the probability that someone who randomly selected one of the airline's flights over the 20-year study period would be killed en route. **based on limited information Sources: Arnold Barnett, MIT. From FAA DATA, International Civil Aviation Organization, Digest of statistics and Flight International Magazine