Boe­ing is haunted by a 50-year-old fea­ture of 737 jets

Los Angeles Times - - FRONT PAGE - By Ralph Vartabe­dian

A set of stairs may have never caused so much trou­ble in an air­craft.

First in­tro­duced in West Ger­many as a short-hop com­muter jet in the early Cold War, the Boe­ing 737-100 had fold­ing metal stairs at­tached to the fuse­lage that pas­sen­gers climbed to board be­fore air­ports had jet­ways. Ground crews hand-lifted heavy lug­gage into the cargo holds in those days, long be­fore mo­tor­ized belt load­ers were widely avail­able.

That low-to-the-ground de­sign was a plus in 1968, but it has proved to be a con­straint that en­gi­neers mod­ern­iz­ing the 737 have had to work around ever since. The com­pro­mises re­quired to push for­ward a more fuel-ef­fi­cient ver­sion of the plane — with larger en­gines and al­tered aero­dy­nam­ics — led to the com­plex flight con­trol soft­ware sys­tem that is now un­der in­ves­ti­ga­tion in two fa­tal crashes over the last five months.

Boe­ing’s prob­lems deep­ened Thurs­day, when the com­pany an­nounced it was stop­ping de­liv­ery of the air­craft af­ter the Fed­eral Avi­a­tion Ad­min­is­tra­tion’s de­ci­sion Wed­nes­day to ground the air­craft.

“We con­tinue to build 737 Max air­planes, while as­sess­ing how the sit­u­a­tion, in­clud­ing po­ten­tial ca­pac­ity

con­straints, will im­pact our pro­duc­tion sys­tem,” the Chicago com­pany said in a state­ment.

The cri­sis comes af­ter 50 years of re­mark­able suc­cess in mak­ing the 737 a prof­itable work­horse. To­day, the aero­space gi­ant has a mas­sive back­log of more than 4,700 or­ders for the jet­liner and its sales ac­count for nearly a third of Boe­ing’s profit.

But the de­ci­sion to con­tinue mod­ern­iz­ing the jet, rather than start­ing at some point with a clean de­sign, re­sulted in en­gi­neer­ing chal­lenges that cre­ated un­fore­seen risks.

“Boe­ing has to sit down and ask it­self how long they can keep up­dat­ing this air­plane,” said Dou­glas Moss, an in­struc­tor at USC’s Viterbi Avi­a­tion Safety and Se­cu­rity Pro­gram, a for­mer United Air­lines cap­tain, an at­tor­ney and a for­mer Air Force test pi­lot. “We are get­ting to the point where legacy fea­tures are such a drag on the air­plane that we have to go to a clean-sheet air­plane.”

Few, if any, com­plex prod­ucts de­signed in the 1960s are still man­u­fac­tured to­day. The IBM 360 main­frame com­puter was put out to pas­ture decades ago. The Apollo space­craft is revered his­tory. The Buick Elec­tra 225 is long gone. And Western Elec­tric dial tele­phones are seen only in clas­sic movies.

To­day’s 737 is a sub­stan­tially dif­fer­ent sys­tem from the orig­i­nal. Boe­ing strength­ened its wings, de­vel­oped new assem­bly tech­nolo­gies and put in mod­ern cock­pit elec­tron­ics. The changes al­lowed the 737 to out­live both the Boe­ing 757 and 767, which were de­vel­oped decades later and then re­tired.

Over the years, the FAA has im­ple­mented new and tougher de­sign re­quire­ments, but a de­riv­a­tive gets many of the de­signs grand­fa­thered in, Moss said.

“It is cheaper and eas­ier to do a de­riv­a­tive than a new air­craft,” said Robert Ditchey, an en­gi­neer, avi­a­tion safety con­sul­tant and founder of Amer­ica West Air­lines, which pur­chased some of the early 737 mod­els. “It is eas­ier to cer­tifi­cate it.”

But some as­pects of the legacy 737 de­sign are vin­tage headaches, such as the ground clear­ance de­signed to al­low a stair­case that’s now ob­so­lete. “They wanted it close to the ground for board­ing,” Ditchey said.

An­drew Skow, founder of Tiger Cen­tury Air­craft, which de­vel­ops cock­pit safety sys­tems, and a for­mer Northrop Grum­man chief en­gi­neer, said Boe­ing has had a good record mod­ern­iz­ing the 737. But he said, “They may have pushed it too far.”

To han­dle a longer fuse­lage and more pas­sen­gers, Boe­ing added larger, more pow­er­ful en­gines, but that re­quired it to re­po­si­tion them to main­tain ground clear­ance. As a re­sult, the 737 can pitch up un­der cer­tain cir­cum­stances. Soft­ware, known as the Ma­neu­ver­ing Char­ac­ter­is­tics Aug­men­ta­tion Sys­tem, was added to coun­ter­act that ten­dency.

It was that soft­ware that is be­lieved to have been in­volved in a Lion Air crash in In­done­sia in Oc­to­ber.

The soft­ware er­ro­neously thought the air­craft was at risk of los­ing lift and stalling — be­cause of a mal­func­tion­ing sen­sor — and or­dered the sta­bi­lizer at the rear to put it into a se­ries of sharp dives that ul­ti­mately caused the plane to crash into the Java Sea.

What hap­pened on the Ethiopian Air­lines flight is less clear, but track­ing data show that it also en­coun­tered sharp changes in its ver­ti­cal ve­loc­ity and at one point in its climb af­ter take­off lost 400 feet of al­ti­tude. The FAA grounded the jet­liner Wed­nes­day, say­ing that new satel­lite data showed the Ethiopian Air­lines flight dynamics were “very close” to those of the Lion Air jet.

Ethiopia sent “black box” record­ing devices re­cov­ered from the crashed jet to France for anal­y­sis, af­ter re­fus­ing to hand them over to U.S. au­thor­i­ties. The U.S. Na­tional Trans­porta­tion Safety Board still plans to send in­ves­ti­ga­tors to France to help its Bureau of In­quiry and Anal­y­sis for Civil Avi­a­tion Safety.

Air­line crashes sel­dom are caused by a sin­gle fac­tor, and the two 737 ac­ci­dents may yet in­volve poor main­te­nance, pi­lot er­rors and in­ad­e­quate train­ing. But it ap­pears in­creas­ingly likely that Boe­ing’s soft­ware sys­tem and the com­pany’s lack of rec­om­men­da­tions for pi­lot train­ing on it may have played an im­por­tant role in the mishaps.

The en­tire need for the soft­ware sys­tem is fun­da­men­tal to the jet’s his­tory.

The bot­tom of the 737’s en­gines are a min­i­mum of 17 inches above the run­way. By com­par­i­son, the Boe­ing 757 has a min­i­mum clear­ance of 29 inches, ac­cord­ing to Boe­ing spec­i­fi­ca­tion books. The newer 787 Dream­liner has 28 inches or 29 inches, de­pend­ing on the en­gine.

The 737 orig­i­nally was equipped with the Pratt & Whit­ney JT-8 se­ries jets, which had an in­ner fan di­am­e­ter of 49.2 inches. “They looked like cigars, long and skinny,” Moss said.

By com­par­i­son, the LEAP-1b en­gines on the Max 8 have a di­am­e­ter of 69 inches, nearly 20 inches more than the orig­i­nal. There wouldn’t be enough clear­ance with­out some kind of mod­i­fi­ca­tion.

In the 737-300, which came af­ter the orig­i­nal planes sold in West Ger­many, Boe­ing came up with an un­usual fix: It cre­ated a flat bot­tom on the na­celle (the shroud around the fan), cre­at­ing what pi­lots came to call the “ham­ster pouch.”

“They made it work,” said Ditchey, whose Amer­ica West was one of the orig­i­nal cus­tomers of the 737-300.

But the LEAP en­gines re­quired an even big­ger change. Boe­ing re­designed the py­lons, the struc­ture that holds the en­gine to the wing, ex­tend­ing them far­ther for­ward and higher up. It gave the needed 17 inches of clear­ance. The com­pany also put in a higher nose land­ing gear.

The change, how­ever, af­fected the plane’s aero­dy­nam­ics. Boe­ing dis­cov­ered the new po­si­tion of the en­gines in­creased the lift of the air­craft, cre­at­ing a ten­dency for the nose to pitch up.

The so­lu­tion was MCAS, which or­dered the sta­bi­lizer to push down the nose if the “an­gle of at­tack” — or an­gle that air flows over the wings — got too high. The MCAS de­pends on data from two sen­sors. But on the Lion Air flight, the MCAS re­lied on a sen­sor that was er­ro­neously re­port­ing a high an­gle of at­tack when the plane was nowhere near a stall.

The pi­lots tried to coun­ter­act the nose-down move­ments by pulling back on the yoke. But even pulling with all their might they could not coun­ter­act the forces, ac­cord­ing to data in a pre­lim­i­nary ac­ci­dent in­ves­ti­ga­tion re­port.

Skow crit­i­cized MCAS, say­ing it acted only on the ba­sis of an­gle of at­tack. The Lion Air jet was trav­el­ing so fast that when MCAS or­dered the sta­bi­lizer to pitch the nose down it was a vi­o­lent re­ac­tion. The soft­ware should have fac­tored in air speed, he said, which would have bet­ter cal­i­brated the pi­lots’ re­ac­tion.

Skow’s firm has de­vel­oped a cock­pit dis­play sys­tem that he says would have iden­ti­fied the fail­ure of the an­gle of at­tack sen­sor and al­lowed the crew to abort the take­off. “We be­lieve we could have pre­vented the ac­ci­dent,” he said.

If the re­sults of the in­ves­ti­ga­tion do not un­der­mine the fun­da­men­tal de­sign of the air­craft, then the 737 Max’s fu­ture may not be in peril, avi­a­tion ex­perts said. It may turn out all that’s needed is a soft­ware fix or ad­di­tional pi­lot train­ing.

The 737 has sur­vived other crises. In a 1988 ac­ci­dent on a flight be­tween Honolulu and Hilo, the en­tire top of the plane came off in an ex­plo­sive de­com­pres­sion. A flight at­ten­dant was sucked out and 65 pas­sen­gers and crew were in­jured. It was blamed on faulty lap joints in the alu­minum skin of the fuse­lage, which Boe­ing reengi­neered.

“The 737 is the most suc­cess­ful com­mer­cial jet ever pro­duced,” said John Cox, an air safety ex­pert and vet­eran pi­lot, adding that com­mon­al­ity among its mod­els helps air­lines with pi­lot train­ing. “It is near­ing the end of its pro­duc­tion life. The tech­nol­ogy will even­tu­ally drive Boe­ing to a re­place­ment.”

Robert Sorbo As­so­ci­ated Press

BOE­ING’S cri­sis comes af­ter 50 years of re­mark­able suc­cess in mak­ing the 737 a prof­itable work­horse. Above, the roll­out of the first 737-700 in 1996 in Ren­ton, Wash.

Ted S. War­ren As­so­ci­ated Press

THE BOE­ING 737’S low-to-the-ground de­sign was a plus in 1968, but it has proved to be a con­straint that en­gi­neers have had to work around ever since.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.