Showing posts with label pilot error. Show all posts
Showing posts with label pilot error. Show all posts

Monday, March 28, 2022

China's Flight 5735 Crash: Mechanical Failure or Human Action?

 

On Monday, March 21, China Eastern Airlines flight 5735 took off from Kunming in western China on a flight to Guangzhou,  on the southern coast.  After takeoff, the plane rose to a cruising altitude of 29,000 feet and leveled off, traveling east.  A little more than an hour into the flight, the plane abruptly took a nosedive and crashed into a muddy mountainous region after falling about five miles in a minute and a half.  Recovery crews at the scene confirmed there were no survivors among the 123 passengers and nine crew members.  The two flight recorders ("black boxes") were recovered and are undergoing analysis, but Chinese officials emphasize that it is too early to reach any conclusions about the cause of this tragedy, which is the worst air transport disaster in China since 2010.

 

As a precaution, China has grounded over 200 of its Boeing 737-800 series airliners for heightened safety inspections.  So far, no other country has taken similar steps, although the Boeing 737-800 is reportedly the most popular airliner of its size class in the world.  One reason may be that the circumstances of the crash do not seem to indicate any particular mechanical failure. 

 

Many types of mechanical problems in aircraft give some sort of warning before causing life-threatening conditions.  For example, the software problems that led to the crashes of two 737 MAX airplanes in 2019 were known to cause unexpected behavior of the aircraft in certain conditions, but the manufacturer assumed (wrongly) that training would enable pilots to take the appropriate actions to counteract the plane's tendency to nose up and stall. 

 

The information we have so far about Flight 5735 shows no sign that anything was wrong prior to the moment that the plane took a nosedive.  Even more oddly, despite repeated attempts by air traffic controllers to contact the pilots during the 90-second dive, no replies were received.  The lack of response would be understandable if an explosion or other violent breakup of the plane occurred in midair, but eyewitnesses and the crash scene testify that the plane was largely intact until it hit the ground, except for a small piece found some 10 km away that might have broken off during the dive.  Investigators estimate that the plane's velocity could have approached the speed of sound, which is beyond its design speed and could have broken small parts off before the crash.

 

Although Chinese officials have not addressed the question directly, one possibility that must be addressed is that one or more of the pilots deliberately crashed the plane.  While very unlikely, such a thing has happened before. 

 

In 2014, Malaysian Airlines Flight 370 disappeared over the Indian Ocean with 239 people on board, and despite two extensive searches, the plane has never been found, although small pieces have washed ashore at various places.  In an extensively researched article in the Atlantic, William Langewiesche makes a strong case for the idea that the pilot, Zaharie Ahmad Shah, somehow overcame the other crew members, depressurized the cabin at an altitude that would have killed everyone on board outside the oxygen-supplied cockpit, and then let the plane crash in the ocean, possibly after committing suicide himself. 

 

It is too early to draw any definite conclusions from the preliminary data we have about Flight 5735, and we may have to wait quite a while before the official investigation produces any results.  By United Nations rules, the country in which an airplane crash occurs is in charge of the crash investigation, and while there are ties to U. S. investigation agencies by virtue of the fact that Boeing made the plane, they must play secondary roles. 

 

Information from the cockpit voice recorders, if any, will be extremely helpful in determining what was going on before and during the last minute and a half of the flight.  Reports state that no conceivable malfunction of the autopilot systems would cause the plane to take and persist in such an extreme nosedive, which apparently can only be induced by the pilot's manual intervention. 

 

This raises a more general question:  should new planes have software in place that would make it impossible to take such actions that would lead inevitably to the loss of the aircraft?  In discussions about how the MCAS system, which is mostly software, led to the crashes of two 737 MAX planes in 2019, it looked like we had already gone too far in the direction of a robot taking over a plane.  In any case, all new airliners are "fly-by-wire" in the sense that the pilot's manipulations of the controls pass through software systems that then execute the actual movement of control surfaces. 

 

In that view, it just becomes a matter of judgment on the part of the computer whether to let what the pilot is telling the computer to do take precedence over what the computer knows is best for the plane, or at least what might lead the plane to crash. 

 

Perhaps the rarity of a pilot going berserk is enough to allow us to lodge our ultimate trust in the person actually flying the plane, and not in some software that probably hasn't been tested under the conditions that would apply if a rogue pilot was determined to crash the plane.  Besides, a pilot that's smart enough to fly a plane safely is probably smart enough to outwit software designed to keep him from doing it. 

 

In any case, we will simply have to wait until Chinese accident-investigation officials reach their considered conclusions about what caused the crash of Flight 5735.  In his article on the Malaysian Air crash, Langewiesche says that important personal information about the pilot Shah was suppressed by accident investigators, who were reluctant to admit the possibility that one of their own pilots deliberately caused the crash, which officially still has no determined cause.  Political considerations can be a factor in accident investigations, and if the Chinese investigation comes up empty-handed, it may be because stating the most likely cause would be politically embarrassing.

 

Politics aside, if a mechanical failure was at fault, it's important to get to the bottom of it so that all the other users of 737-800s can avoid such disasters in the future.

 

Sources:  William Langewiesche's article "What Really Happened to Malaysia's Missing Airplane" is available at https://www.theatlantic.com/magazine/archive/2019/07/mh370-malaysia-airlines/590653/.  I also referred to articles on the Flight 5735 crash at the following sites:  https://www.cnn.com/2022/03/26/china/china-eastern-airlines-crash-intl/index.html, https://www.bbc.com/news/world-asia-china-60887970, https://www.ndtv.com/world-news/china-plane-crash-china-crash-mystery-deepens-as-evidence-suggests-mid-air-breakup-2842114, and https://www.bloomberg.com/news/articles/2022-03-21/china-plane-crash-update-u-s-taps-investigator-boeing-falls. 

Monday, August 05, 2013

The Things That Didn’t Happen To Flight 214


Just moments before Asiana Airlines Flight 214 was to land at the San Francisco International Airport on July 6, some passengers noticed that backdraft from the jet engines was kicking up seawater.  This usually doesn’t happen on normal approaches to Runway 28L, which extends from just behind a seawall that faces San Francisco Bay onto land.  A few seconds later, the main landing gear hit the seawall and sheared off.  After that impact, both engines and the tail section came off, carrying some passengers and crew with it.  The main fuselage slammed into the runway and spun almost completely around before grinding to a halt. 

Flight attendants sprang into action, assisting passengers who needed help in exiting the aircraft.  One injured girl was pulled from the plane by a first responder, only to be covered in firefighting foam from arriving fire trucks.  Sadly, another emergency vehicle’s driver failed to see her underneath the foam, and she was struck and killed.  Another passenger died at the scene and a third passed away a few days later from injuries.  All of the other 304 people aboard survived, including all the pilots and crew, although some sustained serious injuries.  After the plane was evacuated, a fire from an oil leak demolished much of the fuselage, but without injuring anyone.

Any fatal accident involving air travel is a tragedy—usually an avoidable one.  But this accident could have been much worse, and that fact carries with it some implicit good news. 

For one thing, the Boeing 777 involved is a model that was introduced in 1995, and this 2013 accident is the first one involving loss of passenger lives in a flight-related accident.  Although fatal accidents have occurred earlier, they involved refueling or other ground-based situations.  This is an outstanding safety record compared to planes developed during the earlier years of aviation.

Another fact worth noting is that the landing gear was purposely designed to break away under a sufficiently large impact, rather than staying attached to cause a destructive nosedive.  We are familiar with breakaway traffic signs on highways, but I wasn’t aware until now that the same principle has been designed into landing gear.

Finally, the fact that the fuselage endured the abuse of skidding thousands of feet down the runway sans landing gear and kept the remaining fuel from catching fire, staying together long enough for everyone to escape, is a testimonial to its structural engineering.  I am no mechanical engineer, but somebody did something right to make a fuselage that would hang in there during such a trial.

There are things that no airframe can endure, of course.  If the plane had encountered a large immovable object, for example, the outcome might have been quite different.  A similar accident in some ways to the Asiana Airlines crash took place on August 2, 1985.  A Delta Airlines Lockheed L-1011 with 163 people on board was caught in a microburst and windshear during a thunderstorm at the Dallas-Fort Worth Airport during its final landing approach.  The sudden loss of airspeed and accompanying downdraft forced the plane to the ground north of the runway, where it skidded into some giant water tanks and exploded.  Only 26 people survived.  Windshear detectors have since been installed at many airports, and pilots are much more aware of the dangers of such conditions, so the cause of that particular crash is much less likely to occur these days.

The cause of the Asiana crash is still under investigation, but attention has been focused on the flight crew, which consisted of three captains and a first officer.  The man actually flying the plane at the time of the crash had less than fifty hours’ experience on 777s, and was being instructed by the pilot in command, who occupied the co-pilot’s seat at the time.  The runway’s instrument landing system (ILS) vertical glide slope was out of service and a notice had been issued to that effect.  This made it impossible to execute an ILS landing to the runway.  Records indicate that the various automated landing-assistance systems were manipulated during the approach, and it may not have been clear to the flight crew that their approach was too low and slow until it was too late to do anything about it.  The laws of inertia are always in force, and a lot of advance planning has to be done to bring a huge heavy object like a 777 in contact with the ground safely.  Although final conclusions will have to await the completion of the ongoing investigations, it appears that pilot error may be at the bottom of this accident.

As long as human pilots fly planes, we will always have to contend with the possibility of pilot error.  But in general, air travel is safer now than it has ever been, in terms of fatalities per passenger-mile flown.  Even the absolute numbers of fatalities per year, which obviously stood at zero until the invention of the airplane, continues a downward trend that began in the 1970s, and is the lowest since about 1954.  And the total number of passenger-miles flown in 1950 was only about 2% of what it was in 1990. 

The Asiana crash may have stemmed from confusion about who was in charge—the autopilot mechanisms or the real pilot.  But for the vast majority of planes and flights, the amazing system of man and machine called air travel operates efficiently, economically, and with a safety record that was unimaginable in the early days of flight.

Sources:  I referred to the Wikipedia articles on “Asiana Airlines Flight 214,” “Delta Airlines Flight 191,” “USAirways Flight 1549,” and “Aviation safety.”  I also obtained statistics on air travel safety from a paper by Prof. Dan Bogart of UC Irvine which can be found at http://www.socsci.uci.edu/~dbogart/transport_momentusBogart_6.11.12.pdf.