Jet Crash Forum

Newspapers around the world are reporting that the results of autopsies carried out on victims of the Air France disaster sugget the Airbus A330 broke up in mid air.
This would suggest that the weather had a big role to play in this disaster. A force of such magnitude able to destroy an airplane...To all specialists out there. Possible?

This is what I sort of suggested in this post:

http://www.jetcrashforum.com/index.php?topic=45.0

Mid-air break-up definitely happened with AA587 in 2001. The vertical stabiliser snapped off, and one engine also fell off before ground impact. An engine may fall away from the aircraft cleanly, or may hit wing/hull causing extra damage. Engines are fairly heavy - a CF6 family engine weighs about 4.3-4.5 tonnes depending on spec. In 2004 an engine fell off a 747, and in 2007 an engine fell off a 737. In both cases the aircraft landed successfully. So losing an engine - literally - doesn't necessarily cause an aircraft to become unflyable.

Modern jet wings must have a safety factor of 1.5 to be certified - ie they won't break until a load of 50% more than a designer's worst scenario is applied. That's a huge force. To snap the hull in mid-air on a newish aircraft would also require huge force. I've read an anecdotal account from a passenger of a hull appearing to flex visibly in extreme turbulence - although I think that was a 757 or 767.

Could there have been any chance of a major Hail storm cracking or breaking the Cockpit windows?

I know this sounds bizarre, but would it be a possibility?

At least this might explain why there were no Mayday calls or any communication at all from the pilots in those last minutes.

It just seems that for some reason, the pilots were incapacitated.

Any thoughts?

1)Problem of Coffin Corner

2) http://blogs.law.harvard.edu/philg/2009 ... -thoughts/

"How could this happen? Those same pilots would have had unusual attitude training in a Cessna 172 and they did fine. There are a few important differences between a Cessna 172 and an Airbus. The unusual attitude training was 20 minutes into a flight during the daytime. The pilots were prepared for it. It takes a long time to push the Cessna 172 over its speed limit or beyond its stress limits. Pushing the nose down on a jet, by contrast, builds up airspeed at a frightening pace. The Cessna is very tough to spin and can be easily recovered from a spin. A multi-engine jet need not demonstrate spin-resistance or spin recovery. The assumption is that the plane will spend its entire life within a normal envelope of flight attitudes and airspeeds. The Cessna 172 is built to withstand nearly 4Gs and can handle more at the cost of some bending. An airliner is designed to withstand 2.5Gs and the Airbus planes have sometimes had trouble even meeting that standard (if you built an airliner as strong as a four-seat airplane you wouldn’t be able to carry as many passengers)."

Actually, the forces of the storm (wind gusts, etc.) are not necessary to explain mid-air break up.

The previous post refers to 'coffin corner'.  This means that at cruise altitude, if the airspeed gets a little too high or a little too low, the ability of the pilots to control the aircraft is impaired.

When the airspeed indication is correct, this is no problem:  airliners almost never get into trouble from cruise airspeed errors, and if the plane starts to misbehave, a glance at the airspeed indicator is enough for the pilots to know how to correct the situation.

When the airspeed indication is lost, and especially when it is lost suddenly, as apparently happened to AF447, getting the plane safely out of its predicament is (to the extent that I understand the piloting challenges involved) like threading a needle.  If the flight crew was unable to successfully thread this needle, then:

1.  Airspeed Deviation -> Gross Upset

If the speed drifts either too high or too low, the plane becomes unstable, and it could be extremely difficult to return the plane to stable flight without a working airspeed indicator.  The unstable aircraft can easily get into what is called a 'gross upset', basically meaning that it is diving or spinning in an uncontrolled fashion.  The high airspeeds associated with such an event could, by themselves, explain the breakup.  A diving airliner can easily exceed the speed of sound, and its structure is not strong enough to withstand the aerodynamic forces at such speeds.

Probably, modern aircrew are not trained sufficiently if how to recover from a gross upset (rather, they are trained to avoid them).  But even if they are well-trained in recovery, I have studied detailed procedures for upset recovery, written by a test pilot who specialized in the handling qualities of jet airliners -- and these recovery maneuvers depend on a working airspeed indicator.

2.  Airspeed Deviation -> Instability -> Tail Overload -> Unrecoverable Gross Upset

If the speed drifts either too high or too low, the plane becomes unstable, and it could be extremely difficult to return the plane to stable flight without a working airspeed indicator.  If the plane left the narrow range of stable-flight airspeed, and the crew was wrestling to regain control, they may have made large inputs to the rudder pedals.  According to the ACARS messages, the rudder travel limiting system had shut down, because it cannot function correctly without airspeed data.  If the pilots made a large rudder input -- and without knowing their airspeed, they would not have known how much rudder input was safe -- this could break off the vertical tail, making the airplane uncontrollable.

The uncontrollable ship then quite likely would reach an air velocity sufficient to cause breakup.


These are just two scenarios, they are not the only possibilities.  But they could happen in smooth, clear air.  You don't need the storm to explain the breakup, when airspeed data are lost.

That being said, when I write about "threading the needle" to get an airplane safely out of cruise condition without airspeed information, this maneuver could have been made much more difficult by the storm.  In this case, the storm would be an indirect cause, but it would then be misleading to suppose that the "force of the storm" broke up the plane.

First, I think that it is correct that the current evidence suggests (read "leans towards but doesn't prove") a midair breakup.  Most likely this would be the result of an aircraft upset.  What we can rule out is a high-speed, high angle descent into the water.  This leaves a lower-speed lower angle descent as a possibility, or a midair breakup.  The debris field could be explained both ways, as could the injuries.  However....  the lack of a "Mayday" suggests a midair breakup is more likely than a controlled descent.

One other issue.... Without airspeed data, it IS possible to sometimes fly thrust and attitude.  However, without IR, one can't do this.  Note that jet upsets can happen both from unusual bank angles and bad speeds (the Adam Air crash was proximally caused by bad bank angles/bad attitude).  Without attitude data (from the IR units) the plane would have been unflyable in anything other than visual meteorological conditions (i.e. in daylight or at least moonlight with the ability to see the horizon).  When you add bad airspeed data and bad weather data to that equation, you have no hope of recovery.

@einhverfr:

While I agree that in theory, a plane could escape from the "coffin corner" by flying thrust and pitch attitude, I wonder what the probability of success would be.  I hope that an airline pilot will speak to this question.

AF447 may have been cruising in a condition where an airspeed deviation of as little as 15 knots in either direction would have reduced stability of flight.  Even if the pilots had the throttles at the exact setting, and were flying in calm smooth air, suppose they were 0.5 degree off on pitch attitude:  could they have kept the airspeed within limits?  Remember, once the plane started either stall or Mach buffet, the pilots might not have been able to distinguish whether they were too fast, or too slow.  In the case of overspeed, Mach effects could have meant that maintaining the cruise pitch attitude would become the wrong thing to do.

But I presupposed above that the throttles were at the exact required setting.  By the time the autothrottle disconnected, it may have made a substantial deviation from cruise thrust, in response to increasingly unreliable air data measurements.  Unless the pilots had made a mental note of the power settings just before the pitots failed, how would they determine the correct power setting for their present weight, air density, and air pressure?  How many seconds would the crew have needed to compute or look up the required power setting?  And if the airspeed had already deviated (because the air data system was failing) and the throttles were at the wrong setting, how many seconds would they have had to "figure it out", before their plane was in serious trouble?

But up to this point, I ALSO presupposed calm smooth air.  Depending on the degree of turbulence, the flight control system may have been making frequent adjustments to thrust and pitch attitude.

Even if, in the absence of airspeed indication, the pilots had set the exact throttle, and maintained the exact pitch attitude, and kept the wings level, so as to maintain the required airspeed...  AF447 may have been encountering significant gusting, perhaps well in excess of 50 fps.  So even if the pilots were doing everything humanly possible, as quickly as possible, to fly without airspeed indication, mightn't the plane have fallen into stall, or overspeed?

I'm just a humble engineer (not specialized in aviation), and only an "armchair pilot."  So maybe my analysis is wrong here (anyone, please criticize!).  But unless I'm missing something, an abrupt loss of airspeed indication at cruise is an acutely dangerous crisis, and this danger would be magnified by a storm environment.

Hoping an airline pilot will talk to us about their training for this situation....

AVBuff:

My point is that airspeed info is extremely important.  However, inertial reference is even MORE important, and the loss of both would be catastrophic in a cloud at night.  If one had to choose one or the other, attitude reference will be the more important one.

I agree, loss of attitude instruments at night in foul weather is unsurvivable, period.

But I suspect that even if the ADI continued to function perfectly, AF447 may have been doomed.  Unless there is a mistake in the analysis I presented above, no amount of skill would have sufficed to save the plane:  it would have also needed some luck to stay within the narrow safe airspeed range, without any feedback to correct airspeed drift that would result from tiny control errors, or the stormy air mass.

I was a jet fighter pilot in the Royal Canadian Air Force some years ago. My rather dated experience prompts me to ask if modern jet airliners have artificial horizons, turn-and-bank indicators, and altimeters which are not "tied" to the electronic systems which have received so much attention in connection with AF 447's unfortunate demise? If not, sophisticated as these newer systems may be, it seems to me a pity (at least) that there is no capability for the "stick-and-rudder" guy to get out of a tight spot when they argue with each other and turn off (apparently).

One further question: it would be interesting to learn from knowledgeable source(s) how long it would take for the ice blockage of the pitot tube(s) to occur? If it were at all gradual, would this not be noticed by the electronics and/or the aircrew?

Extrapolating from AF447, but not necessarily talking about AF447:

—Supposing (in a future situation) that an aircraft finds itself at night, over the Atlantic, in a tower of cumulonimbus.

—Supposing the Pitot tubes show contradictory speeds, because of freezing or for any other reason that is possible.

—Instead of the autopilot coming on in such nearly hopeless conditions, wouldn't it be better to do what any kid with a computer routinely does?

I relate this to a kid at a computer because very often, in the relative safety of his suburban bedroom, the young computer-whiz simply goes straight ahead on a "What IF" assumption, taking his computer a few keystrokes closer to a virus or a crash but with an overview of relative known danger, including his ability to contend with a virus that might be at hand. When I learned computers, this "dangerous thinking" was new to my scientific brain, and I had to get used to it. In pure logic, thinking is zero-based, and we don't proceed without reason. In the midst of a thunderhead at night over the Atlantic, we can't afford to be so pure. The pilots were being handed a superhuman challenge, and we seem agreed that they were very very unlikely to be able to contend with unknown speed in a coffin corner. Would five seconds' more time to think be of help in a future situation?

Backed way into coffin corner, with no time to think, wouldn't the pilots be best to have the airplane attempt to fly at NEARLY top possible speed, without actually hitting the speed at which engineers would reckon it would likely disintegrate?

—In such an emergency-only situation, the pilots might gain a few seconds in which to attempt something else. Maybe they would fail, and the aircraft would be lost. BUT, THE DIFFERENCE IS that the plane would have a "plan" of a duration of just a few seconds. This might be better than the pilots' being handed their autopilot's "Man, this is rough—I'm outa here" reaction which was apparently what happened.

My own opinion is that the AF447 pilots attempted to fly too fast, and that resulted in the disintegration of the aircraft (using "disintegration" to mean dis-integrated-ness).

But perhaps a different device, not a Pitot tube–or else a computer calculation based on pre-programmed probability in case of Pitot tube disagreement—could set a plane's attempted speed for a vital few seconds. It might have to rely on software making a "best guess". But it might reduce a 99% chance of disaster to a 90% chance of disaster—and that would be a big improvement.

@Alouette

You wrote:

"My rather dated experience prompts me to ask if modern jet airliners have artificial horizons, turn-and-bank indicators, and altimeters which are not "tied" to the electronic systems which have received so much attention in connection with AF 447's unfortunate demise? If not, sophisticated as these newer systems may be, it seems to me a pity (at least) that there is no capability for the "stick-and-rudder" guy to get out of a tight spot when they argue with each other and turn off (apparently)."

I agree with you.Do they also have an information about the load of G-factor?Do they know the limit of G-load of their plane? Pilots today do learn less. No experience in glider-flying (its important for the feeling for an aircraft in turbulence because gliders are seeking turbulence to climb up), no experience in acrobatic flight (to handle a plane under extremly circumstances), no experience of jetfighter
flying (maneuver in speed,altitude....) and so on.
Flying only by computer and autopilot is to less.

The problem of Coffin Corner is higher at equator-thunderstorms if you fly from Brazil to Europe than from Europe to Brazil because of the weight.

My conclusion is:

It is impossible for this kind of pilots to fly consciously into unknown (Top,turbulence,cells....) thunderstorms at the Coffin Corner.

Broken in mid air ?

What is the limit of G-load of an Airbus ?

Generally, Part 121 aircraft are certified in the structural load range from -1.0 G to +2.5 G in their clean (high-lift systems retracted) configuration.

Aerodynamically, such aircraft can be capable of generating loads considerably in excess of these limits (for example, when attempting to recover from a dive at low altitude).

I don't know how deeply airframe stresses have been studied in the situation where an airliner goes supersonic in a dive.

I also don't know what the history is of modern jets breaking into large sections (excepting vertical stabilizer separation) solely due to aerodynamic forces - this may be so rare that there are not examples to study.