Pageviews last month

Wednesday, 3 June 2009

Flight 447: Answering the Questions

As anyone who was following the news on June 1, 2009 knows, Air France Flight 447 turned up missing in the mid-Atlantic en route from Rio to Paris. I'm opening up this post now, on June 3, as reports come in that the crash site has been located and recovery units are headed to the scene. I'll continue to visit this situation and add more information as the situation develops. I intend to address the following questions:

1) Is this the greatest air disaster in history?

2) Could there be any survivors?

3) Could it have been a terrorist attack?

4) Could it have been caused by bad weather?

5) Will the cockpit data recorders give any clues as to the cause of the crash?

First, I'd like to comment on the phrase now being used, "Lost over the Atlantic."

Despite the settled claim the wording has in the English language, Flight 447 was never "lost over the Atlantic." This well-turned phrase hearkens back to the early days of Lindbergh and Balchen, when planes took off from one coast, not to be heard from until they reached the other. If they didn't show up, they were "lost over the Atlantic," which could mean that they got off course and landed somewhere else, ran out of fuel and had to ditch, or perhaps had actually flown into the ocean in bad weather.

None of these things happen now. Both pilots and air controllers know within a few miles where a commercial airplane is at any moment. Yes, pilots sometimes divert from their course to get around a thunderstorm, but they always let the controllers know what they are doing. A quarter of an hour before Flight 447 broke up in midair, the pilot sent a signal back to Brazil that he was flying into a thundercloud. It was clear after the burst of automatic signals had reached Paris that Flight 447 was lost--not over, but in, the Atlantic. And it is in the Atlantic that searchers have been looking for the wreckage; first on the surface, and eventually at the bottom.

Now on to Question Number One.

Although this is the worst air disaster in Air France's history, there have been several airplane crashes with far greater loss of life, beginning with the Tenerife disaster, when KLM Flight 4805 flew into Pan Am Flight 1736 on takeoff, killing all 248 aboard the KLM 747 and 335 on the Pan Am plane. The deaths can be indirectly blamed on a terrorist attack; both planes had landed at the inadequate Tenerife runway after a small explosion in the terminal of the Las Palmas airport rendered it temporarily off-limits to incoming flights. Most of the other nine reasons for the crash were due to technology limitations and the human factor, along with a fear of violating Dutch overtime laws.

The second worse airplane crash ever took 520 lives aboard Japan Air Lines Flight 123 on August 12, 1985. This was, at the time, the deadliest crash ever, and the number of fatalities in a single airplane has never been exceeded, and probably never will be; few planes ever take off that heavily loaded. The 747 on a domestic run had just reached cruising altitude after takeoff from Tokyo when the rupture of a faultily repaired aft bulkhead, damaged in a hard landing seven years earlier, drained the hydraulic system, rendering the jet almost uncontrollable. After a 32-minute struggle to gain control of the plane, the flight crew brought it down so low that it hit one of the foothills of Mt. Fuji. The maintenance supervisor responsible for signing off on the plane's airworthiness later committed suicide--the 521st fatality from the crash.

What's amazing is that even though the airplane clipped the side of a mountain, flipped over, and landed on its back at hundreds of miles an hour, not everyone died. It's unknown for sure how many survived the impact--one survivor landed in a tree--but by the time rescuers finally got around to making their way to the crash site, they only found four survivors, and evidence that many more had died of their injuries overnight while awaiting rescue.

The single worst terrorist-caused air disaster was on June 23, 1985, when a suitcase bomb blew up Air India Flight 182 between Montreal and London. The 747 fell out of the sky into 2,000 meters of water 120 miles off the Irish coast, killing all 329 aboard--more than died aboard all three hijacked airplanes on 9/11. Subsequent medical examination of the bodies showed that some had died of drowning, indicating that they were still breathing when their bodies hit the water after a drop of 31,000 feet.

So the first question is not hard to answer--this isn't the worst air disaster in history; it doesn't even make the top ten. It's not the most people to die when an airplane broke up on or over the ocean; it's not even the most for a 2-engine airplane, although all but one of the top 10 (an Airbus 300 shot down over the Persian Gulf in 1988) involved 3- or 4-engine aircraft.

This will amaze most of my readers, but they probably haven't even heard about any of the last ten major air disasters, in each of which an average of 163 people died--more than three times the combined air and ground casualties of Continental Connection Flight 3407. Count on it, unless at least one American citizen is killed in an air crash--any air crash--it generally doesn't even rate Page Two coverage.

Moving on to Question Number Two, we can learn some lessons from the worst air disasters: first of all, that an airplane certified as airworthy can carry a hidden factor just waiting to blow it out of the sky--whether that factor be a bomb or a bungled repair job. Secondly, that it's possible to survive being blown out of the sky, but continued survival is contingent upon quick rescue. This is why forces from three countries converged on the area so quickly on June first. But by the time any of them reached the scene, it would have already been too late. Anyone still alive after the plane broke apart was most likely dead by the time the sun rose on Flight 447's fuel slick.

Here's what the people on the plane had to survive, just to reach the storm-tossed surface of the mid-Atlantic [but see second update]:

1) Rapid decompression at 35,000 feet would have popped numerous blood vessels in their eyes and noses. Their hands would have swollen so fast, it's unlikely anyone could have released a seatbelt until they reached lower altitudes. And the captain would certainly have required everyone to be buckled before he took the plane into a thundercloud.
2) Temperatures about 40 degrees below zero would have caused almost instant frostbite.
3) The shock of deceleration from 520 miles an hour would have broken limbs like matchsticks.
4) Pieces of wreckage would have filled the air, with potential extreme trauma injuries.

Even if they survived the shock of hitting the water, they would have penetrated so deeply before resurfacing that their traumatized lungs would probably have filled with water.

There is one very real possibility, though, that the tremendous updrafts associated with the thundercloud that brought down Flight 447 may have contributed to the temporary survival of some passengers. In 1971, several passengers survived the crash of LANSA Airlines Flight 508 when it blew apart in a thunderstorm over the Peruvian jungle. They were still buckled in their seats when they hit the ground, and it is thought that their fall was greatly slowed by updrafts. We'll come back to Flight 508 later.

So if any passengers of Flight 447 did survive the breakup, they would have to have remained buckled in an intact portion of the aircraft, thus being sheltered from the wind and cold as their piece of the airplane descended; assumed a braced position before hitting the water; and managed to swim to the surface before their seat sank to the bottom. It's almost inconceivable, though, that they didn't have enough injuries by then to attract sharks to their bleeding bodies before daybreak brought with it some hope of rescue.

Moving on to the question of a terrorist attack: I can guarantee that France's investigation will conclude that anything else caused Flight 447 to fall out of the sky before they admit that a terrorist bomb did it. And I won't say why. Just oogle the following search terms and you may come to the same conclusion:
Flight 800 shot
Flight 587 flat

But regardless of what the official report ends up saying, I don't believe that a terrorist bomb brought down Flight 447. For one thing, there's a reason why this is Air France's worst air disaster ever: terrorists don't target French airlines. Even Richard Reid, flying out of Paris, chose an American Airlines flight on which to detonate his shoe bomb. Terrorists target airlines like American, TWA, and Pan Am (the last two of which having gone out of business in the wake of their air disasters). A flight from Rio to Paris just wouldn't be on anyone's target list.

Secondly, the timing was too coincidental. A bomb does not just happen to go off--or a meteor strike--just after a plane enters an area of tremendous turbulence and electrical activity. Wings do fall off, and fly-by-wire systems do fail, however, in exactly such circumstances.

I'll answer Question Number Five now, as Question Number Four will take a while.

No doubt the cockpit data recorders will give some clues as to the cause of the crash, but we already have most of the information needed to find out what caused the crash, and it's unlikely that the data recorders will tell us much we don't already know. Voice recordings have shown to be notoriously deficient in explaining what went on in the final seconds of a flight that ended in disaster; you won't generally hear things like, "Oh, there goes the right wing" or "why isn't this @#% thing working?" Pilots in an air disaster are concentrating on keeping the plane under control, not on describing what's going wrong.

The Data Recorder won't tell us how fast the updraft was; how much the wind shear was; whether it was a negative or positive lightning strike; how many gallons per minute of water the engines were taking in; or even how fast the plane was flying, inasmuch as the information we already have gives two conflicting readings.

And this leads us right into the Fourth Question.

You see, part of the contributing factors to this crash, as well as many or most others, was a strange combination of old and new technology. Airplane designers have to decide whether they want the pilot or the computer to fly the plane, because a disagreement between the two often ends in disaster. The aircraft used in Flight 447 was an Airbus A330, which employs a state-of-the-art computer system that is capable of taking over from the pilot when it perceives he isn't flying the plane correctly. Only when all else fails does it turn control back over to the pilot. Yet the instrument feeding airspeed information into the computer was a Pitot tube, an invention that pre-dates the airplane itself and has remained relatively unchanged for over 150 years. The only modification of note is that after it was found to ice up at only a few thousand feet of elevation, an electric heater was added to keep its intake clear of ice. This small heater seems to have proven inadequate to handle the extreme icing conditions possible in a mid-Atlantic tropical thundercloud. Despite this, pilots are given no training on how to recognize, much less respond to, this deadly condition.

As well as the Global Positioning System technology has proved at identifying the location of an aircraft in mid ocean, it has not been approved for use in identifying the rate of change in location, commonly referred to as speed. And had the A330's radar screen been allowed to instead display an online satellite image, it's highly unlikely that the flight crew would have flown into the 100 mph updrafts that were only visible on the latter. Once again, state-of-the-art flight systems are shown to have limitations that are as much legislative as technological in nature.

June 8, 2009
I'm closing out this post today. Many bodies have been recovered, at tremendous expense. Per capita, I don't doubt the three governments will have spent more than the US did on weregild for 9/11 by the time this is all over. Obviously changes in procedures for transoceanic flights will result in the aftermath; see the link in the post title for some suggestions.

The fact that the bodies were found 45 miles farther down the track suggests that the computers on board the Airbus had pushed the craft to an unsustainable speed following faulty input from the frozen pitot tubes. For sure this will be changed, as this has happened before and will happen again if modern technology is not brought to bear on a 200 year old invention.

Having addressed these questions to the best of my ability, I will now leave the discussion to others.

And now I can start posting about something else again.

Well, it's been 10 months since the cockpit and data recorders were found, but the official report still hasn't been issued. Suffice it to say that there won't be a whole lot to add once it is. Flight 447 really was flown into the ocean, for the simple reason that the pilot kept the plane nose-up, at too high of an angle to sustain altitude, from 38,000 feet all the way down to sea level. He did this because the planes' computer was programmed in such a way that every time he lowered the nose, a warning sounded alerting him that the nose was too high. But when he raised the nose again, the alarm went silent. The flight crew, all during their 3 minutes of rapid descent, were being bombarded with frequent and often conflicting computer-generated warnings. The most effective way of dealing with them all appeared to be exactly what was done: flying the aircraft into the sea.

It has been said that all safety regulations are written in blood--the blood of those whose deaths showed the need for them. Inasmuch as aircraft engineers will now take a more human-centered approach to how a plane is flown in crisis situations, they both ensure that the passengers of Flight 477 did not die in vain, and show once more that in a crisis with massive equipment failure, no control system designed by the human brain has ever managed to improve on the human brain itself.

Looking back over this post, I see that I had given a scenario in which the plane fell apart at altitude. This clearly did not happen; passengers would have ridden in relative comfort all the way to the water's surface. However, the scenario in the cockpit is chilling in its resemblance to what happened in the final minute of Flight 3407's approach to Buffalo: amidst much confusion, as a stall warning horn broke the pilot out of his autopilot-induced lassitude, he gave the opposite control input needed to keep the plane in the air. Tragically, at the time Flight 447 went down, the "Lessons Learned" from Flight 3407 were yet to be published. The changes in aviation policy that have resulted from these closely related disasters could mean that from now on, pilots will finally be able to resist the urge to pull up on the stick when in a free fall at low airspeed due to icing.

THIRD UPDATE August 2013
I see that I was going to say more on the crash of Flight 508. I can't remember now what it was, although I have since learned that it set the task of Bible Translation in the Western Amazon Basin back by several months or years. Finally, it appears that the only thing that saved any of the passengers and crew on Asiana Airlines Flight 214 was a willingness of the pilot to take over from the computer before it flew the airplane into the ground. Unfortunately for those who were hurt in the crash, he was too late in taking over to completely prevent the plane from hitting a sea wall at the end of the runway.

For those searching on information regarding Malaysian Airlines Flight 370, follow the links from this update.

FIFTH UPDATE December 2014
Whenever there is an air disaster, this page draws a lot of viewers. So I suppose I should briefly say something about Asia Air Flight 8501.
As with all breaking news stories, we are being fed a lot of disinformation. There were 40 bodies found. No, seven. No, six. Well, two were brought back to land. One was wearing a life jacket. No, it wasn't. And so on. About all I can say with confidence at this point, two days after the crash is:
1. It was weather-related. The plane, like Flight 447, was flying into a thunderstorm.
2. I'm sure that the altimeter problem on this Airbus had been fixed, but there was still the computer vs. pilot problem, and this being a cut-rate airline, I'm afraid the pilot wasn't up to the challenge.
3. There were no survivors.

Among the interesting bits of trivia, this flight having originated in an officially Muslim country:
1. Most of the passengers were Christian Indonesians.
2. Among the passengers was a family of Korean missionaries to Indonesia.
3. Despite there being no Americans on board, this disaster made front-page news in America.

SIXTH UPDATE May 21, 2016
With the demise of Egyptair Flight 804, we are getting used to the scenario of conflicting reports emerging from a search-and-rescue mission. What I should point out in relation to this is that search teams ALWAYS find debris in the area of a downed aircraft, for the simple reason that the ocean is full of it. Initial reports of crash-related debris are therefore almost always followed up by a retraction. 

SEVENTH UPDATE: an historical account of the crash

No comments:

Post a Comment

One comment per viewer, please--unless participating in a dialogue.