The 'Coffin Corner' and a 'Mesoscale' Maw


The A330 that crashed a week ago - from

The Air France 447 mystery may never be solved beyond a shadow of doubt, but there are some telling, tragic clues to consider based on what we know about the airplane systems and the extreme weather and aerodynamic conditions it encountered before it went down a week ago.

First, a bit of aerodynamics: The doomed Airbus A-330-200 was flying ever so close to its maximum altitude – in a zone pilots call the “Coffin Corner”. It refers to the edge of so-called “flight envelope” of an aircraft. At this altitude, the air is much thinner and that significantly narrows the swath of speed at which the airplane can safely operate.

Because there are relatively few air molecules passing over the wings, they need to be moving faster to generate enough lift to keep the plane at altitude. They will stop flying (stall) at a much higher speed (true airspeed) than they would on approach to an airport at sea level.

At the other end of the safe speed spectrum is the sound barrier. The wings on an airliner like the A-330 are not designed to break the speed of sound. Venture toward Chuck Yeager country and an airliner will begin buffeting. And as altitude increases, the buffet speed (the sound barrier) decreases (once again the dearth of air molecules is to blame).

So you see the squeeze play as a plane flies toward the Coffin Corner: the margin between the between the high and low speed limits gets thinner and thinner (along with the air).
Matter of fact, given its estimated weight, altitude and the outside air temperature (which also affects air density), AF 447 was flying through the eye of a speed needle only about 25 knots (28 mph) wide.

And one more important point: as jet engines fly higher, they steadily lose their oomph (you know, thin air). Matter of fact, the maximum altitude a plane can safely fly is partially determined by the point where the engines can no longer maintain a minimum rate of climb. In other words, you are supposed to level off just before they go into “Scottie” mode (“No more power, Captain!”).

So while you are napping, eating or watching a movie on that flight to LAX, you should know the plane you are flying is cruising along at the ratty edge of its capabilities. Why? Money. The higher an airliner flies, the better gas mileage it gets.

But rest easy, white-knucklers; flying in the “Corner” is routine and safe – so long as the weather is benign, the air is smooth and the sensors, avionics, computers and autopilot are all doing their job.

But of course that was not the case for Air France 447.

The weather where and when the plane went down was horrible – the storms among the meanest weapons in Mother Nature’s arsenal. On their nose, the crew would no doubt have seen the outline of a towering wall of cumulonimbus clouds – illuminated in strobe-light fashion by lightning.

Seeing this would not have been a big surprise to them. Their pre-flight weather briefing would have included satellite imagery clearly depicting the “Cb’s” – as pilots call them. Besides, big storms are a common occurrence over the Atlantic along the equator – where the airflows of two hemispheres collide.


Satellite Image of MCS that Air France 447 flew into - From Tim Vasquez

Meteorologists call the storm AF 447 flew into a “Mesoscale Convective System” – a large complex of multiple thunderstorms where the sum total is greater than the individual parts. MCS storms are obviously much bigger than your garden-variety thunder-bumper – and they last a lot longer. [More about the weather at Tim Vasquez’ insightful, detailed blog.]

Precisely because big thunderstorms are common there, airliners are constantly threading their way through the nastiest cells – deviating at the pilot’s discretion. But no professional pilot would knowingly auger into the heart of a thunderstorm this potent. A pro knows no airliner is designed to survive those conditions – no matter how advanced it is technologically and structurally.

Hard to believe in this day and age, but when you are flying over the pond, you are pretty much on your own. You are not talking to air traffic controllers or being painted by their radar – and of course there are no weather reporting stations beneath you. By definition, thunderstorms are unstable, dynamic and fast-moving. So by the time they reached the storms – more than four hours into the flight – what they learned in the pre-flight briefing was yesterday’s news.


Weather Radar - from NASA

As a result, flight crews rely heavily on the weather radar bolted onto the nose of the airplane. It is a very useful safety device but interpreting its display is a bit of a black art. A lot of pilots, frankly, do not fully understand the intricacies of its capabilities and limitations. It is akin to a blind man with a cane; he can tell something is in his way, but he doesn’t see it.

For instance, the radar mostly detects rain and hail – and if that first layer of storm cells was particularly heavy, it might have acted like a curtain – hiding the reinforcements from radar beams. With the benefit of hindsight (and satellite imagery captured at the time of the crash), we know now there were at four more layers of strong storms behind the first line of cells. And radar cannot detect the strong updrafts of warm air that feed a thunderstorm.


Infrared image of storm - from Tim Vasquez

Did the Air France crew spot a gap in that first line of storms that turned out to be a “sucker hole” – sending them into a box canyon of violent storm cells? Maybe. If they could have seen the full depth and intensity of those storms, would they have changed course to avoid it? Hard to imagine they would say, “Steady as she goes…”

No matter how they made their decision to fly into the maw, it was likely not long before they would have known they made a big mistake.

The last message from the crew – a text – indicated they were flying through thunderstorms with “fortes turbulences” (strong turbulence). Hard to know exactly what he meant. In the US, we define turbulence as “Light”, “Moderate”, “Severe” or “Extreme”.  The FAA defines the latter as “turbulence in which the aircraft is violently tossed about and is practically impossible to control. It may cause structural damage.”

Extreme turbulence is precisely what an airplane would be apt to encounter inside an MCS.

Flight Profile created by Tim Vasquez

Flight Profile created by Tim Vasquez

So why wouldn’t they just make a speedy U-turn at that point? They might have, but attempting a maneuver like that in severe or extreme turbulence would likely have made things worse. Remember, the engines were close to maxed out and the speed margin was minuscule before the plane pierced the storm clouds. Simply banking the wings could be enough to trigger an aerodynamic stall.

And consider this: those updrafts bring warm moist air to higher altitudes – feeding the storm. That also might have increased the air temperature where the Airbus was flying. Warmer air is less dense – with fewer molecules – meaning the airliner might have suddenly been flying above its maximum safe altitude.

Bottom line: if all sensors and systems on the Airbus kept working the aircraft might have been hard pressed to stay aloft. But of course, the systems started crashing before the airplane did.

So were those failures contributing causes of the crash – or simply the upshot of an aircraft taking a beating it could not withstand? Maybe it is a little bit of both.

More on that tomorrow…


Tags: , , , , , , ,

72 Responses to “The 'Coffin Corner' and a 'Mesoscale' Maw”

  1. harwich02 Says:


    I’m glad I found you! I always tuned-in to CNN to get your views on any air incidents as you seem like the best, most articulate source for information. I look forward to your upcoming reports! Keep up the good work!

  2. pixelsmart Says:

    Hi Miles,

    That was a chilling read.

    You did not discuss the pitot tubes and the reports about issues with the air speed indicators. Throwing that into the mix of the above described situation seems like it could push such an already precarious position, to the breaking point.


  3. amy5467 Says:

    Great piece Miles. Thanks for putting the aerodynamics into language that those of us with minds challenged in that area could understand. I think I will look at my overseas flights a bit differently now. So much for the relaxing snooze across the pond. My doctor will need to give me more meds for those trips! I look forward to your next article on the subject.


  4. Gary Andrew Poole Says:


    I am really enjoying your posts on Flight 447. Keep up the good work.


  5. paulgauthier Says:

    Just found your excellent article after Howard Kurtz mentioned this new site. Such a relief to see an informative piece about the Air France crash after suffering through all the detail-free copy-cat articles written by the mainstream media. They all seem written for people reading at the 4th grade level. Great work.

  6. patimc Says:

    Hello Miles,

    To imagine all the facts and knowledge that must be swimming around in that superior brain of yours has given me a headache! Very impressive.

    One teeny bone to pick…you have now made a seasoned flyer mortified! “The plane you are flying is cruising along at the ratty edge of its capabilities”???? Uh…yikes?!? Yes, I understand that you went on to clarify that, but nevertheless, I shall not be able to remove that thought from my mind for a while. *shudders*

    All of the information that I have been reading on the Q Corner parallels your thoughts excatly. Thanks for making it even clearer and for putting more pieces of a huge puzzle in place.

    Very much look forward to more. You are the best.

    • Miles O'Brien Says:

      Why worry? – you are on the edge every day – driving down a narrow highway – inches form oncoming traffic. Sometimes ignorance is bliss!

      • patimc Says:

        Okay that is a good point, I’ll give you that. A wise man once said “Every day is an adventure” and should I have my druthers I would truly rather fly than drive. But if the tail fell off of my car, at least I could get out and walk away…hopefully. 🙂

  7. wordmanx Says:

    Very enlightening and educational, thank you. However, I am driven to ask … just what alternative actions could the pilot(s) have taken that actually might have worked ?? Fate and/or inevitability don’t seem very comforting. Jeff A

    • Miles O'Brien Says:

      They could have opted to deviate around the storm. It might have forced them to make a fuel stop – which airlines/flight crews try to avoid – as they are expensive.

  8. texfountain Says:


    Good article on AF 447.

    Any thoughts on why the rudder limit fault message came so soon. To me, it seems the aircraft had a sudden yaw that caused the engine thrust to be different. That knocked the auto-thrust and autopilot off-line and some ice on the aircraft skin may have moved and blocked a pitot tube(s).

    This website has more on the subject…

    • Miles O'Brien Says:

      I think we need to pay close attention to the rudder. Its motion is limited – based on airspeed – but if it did not know how fast the plane was going – the crew could have easily overstressed it – ala AA587.

      • djini Says:

        Miles –

        Thank you for highlighting AA587’s demise and the bigger issue of laminated rudder designs on the Airbus. Could the erroneous airspeed messages have been caused by a broken off tail?

        Are we the only ones that have linked the two accidents?

      • Miles O'Brien Says:

        More like the bad airspeed data led to an overspeed situation – coupled with a relaxed flight “law” – making it possible to break off the tail.

      • timc Says:

        This is a red herring. The rudder travel limiter failure is part of the normal and well-understood (by Airbus pilots) process of degradation of the flight control systems, as reported by the aircraft via ACARS. The rudder is not used in normal flight in a swept-wing aircraft; its main use is to mitigate the assymetric forces on the aircraft in the event of an engine failure. A cursory read of the AA587 accident report will show that the two accidents are not related. Finally, an examination of the photographs of the recovered portion of the vertical tail would appear to show that the rudder was still attached and was broken in line with the fin, suggesting that the two items were broken simultaneously on inpact with the sea or other solid object, and not by aerodynamic stress to the structure.

      • Miles O'Brien Says:

        How can you say the two accidents are not related – when we do not know what happened to AF 447? And the NTSB says the VS/Rudder broke off because of aerodynamic stress. How do you explain that they were found separate from the debris – first in line?

      • timc Says:

        Sorry Miles, I’ve been busy in my normal jobs of flying Airbus aircraft and raising a family, and couldn’t get back till now.

        The focus on the rudder travel limiter is a red herring because it fails in the status quo if the flight control system degrades for whatever reason. The faster the aircraft is flying, the smaller the rudder travel limit. The ‘failure’ warning does not mean that the available travel is no longer limited; it means the limit is now fixed. If the failure happens at low speed, the now-fixed available travel is large and the airplane warns the pilot to be careful with rudder handling as he could overstress the vertical fin as the leverage capable of being applied increases as the aircraft accelerates. At high speed, such as in the cruise, the available rudder travel is low, and overstress due to careless rudder handling should not be possible. In the AA587 case, Airbus contends that the pilot mishandled the rudder. The issue in that case was, I believe, that the pilot (or a design failure) may have somehow overridden the rudder travel limiter allowing him to apply more rudder than the structure could cope with.

        In the AF447 case, it is, in my opinion, highly likely that the majority of the warnings ACARS reported back to AF HQ were triggered by the break up of the structure of the aircraft due to encountering extreme weather. I suspect that few, if any, of the reported failures are directly causal to the loss of the aircraft and are in fact simply the death throes of a crippled machine. It may well be, however, that the pitot tube problem (causing confusing and unreliable speed indications) hastened the loss of control and subsequent break-up of the aircraft, and that’s where I’d be looking most closely for any weaknesses in the aircraft. However, none of that is as important as why the aircraft was in the weather in the first place. The strongest machine will break if it’s put in the wrong environment; you don’t blame a car for killing its passengers if the driver drives it into a brick wall at 100mph! You may learn plenty about survivability and structural integrity from such an incident, and may apply those lessons to future vehicles, but the cause remains that the driver did the wrong thing.
        In this case, it would appear that the aircraft was in a place it shouldn’t have been. the important question is not what happened inside that weather, but why was it there?

  9. wheel12 Says:

    Would this tragic event be called a series of human failures from flight planning to cockpit decisions with the broad area of thunderstorms in front of the flight?

    Who would make the decision to fly knowing the thunderstorms in front of the flight?

  10. dgoldman Says:

    Given your excellent description, why wouldn’t it be standard practice when flying into areas of large-scale thunderstorms to be at a lower altitude, and thus have a greater margin of error?

  11. johnniepal Says:

    Avoidance IS best. However when operating in a non radar environment it is a long process to get permission to deviate from your course. The request is made via HF radio to a radio operator IF no one else is talking. That request goes to the controller. He makes the decision. It then gets sent back to the radio operator who then radios the decision back to the flight IF/When he is not talking to another flight. Given the work load of the radio operator and the controller it can take some time. Initially the weather could have looked OK. Storms develop rapidly. In the day you can watch them billow up. This may have been the case. They may have seen a hole but got to close as it developed and determined there was not enough time or room to either deviate or turn around. You can’t turn on a dime at 600 Kts.

  12. Doomed AF 447 flight crew drawn into a “sucker’s hole” of towering thunderstorms four layers deep « mysticalmaven Says:

    […] First, a bit of aerodynamics: The doomed Airbus A-330-200 was flying ever so close to its maximum altitude – in a zone pilots call the “Coffin Corner”. It refers to the edge of so-called “flight envelope” of an aircraft. At this altitude, the air is much thinner and that significantly narrows the swath of speed at which the airplane can safely operate. Continue reading at LINK. […]

  13. hidflect Says:

    Wow… Miles writes with real verve and interest here. It’s so different from how I saw his manner on TV that I must figure his natural style while being an anchor on CNN was stepped-on and crimped by the network. In my mind he’s gone from being just another “face” on TV to now being a real-live news engager.

    • Miles O'Brien Says:

      Maybe if I was more engaging on TV, I’d still be there!

      • pixelsmart Says:


        Your quality of work on air was totally up to professional standards. Your exit was not in your hands. CNN over the last decade has become populist and dumbed down. They are a huge disappointment. Enlightened sc & tech reporting just doesn’t fit their agenda.


      • Miles O'Brien Says:

        So be it. I am happy to be where I am. Thanks.

  14. mikezuhl Says:

    The flight envelope graph in the article does not appear to be for an A330. It has a max altitude of roughly 75,000 ft and a max airspeed of roughly 150 knots, while an A330 cruises at 470 knots at 35,000 ft.

    • Miles O'Brien Says:

      I could not find one for the A330 – was showing it as an example of “the envelope”

      • mikezuhl Says:

        I have seen references to the coffin corner with regard to AF 447 but I haven’t been able to find the A330 flight envelope graph or any other raw data. I’m curious as to where you found the “25 knot window.”

        BTW, I was a big fan of your CNN work. Finding you on T/S lead me to sign up and follow you. Keep it up!

      • Miles O'Brien Says:

        great news. thanks.

  15. drsacha Says:


    Your post gives a great sense of adventure we can have flying, will be even better now knowing the danger are far closer flying to space on shuttle. Even if it is not the plasma floating off the window and just ice flakes forming between the window glasses.

    Most of the town bus are tracked with GPS devices and have satellite signal to show some news or songs in between advertisements. Even seeing this around few buses in India also (the bus conductor here feel it is install to monitor if they are cheating hence they bang it against something so that the device stops working 🙂 ). Monopoly of the two companies and the super power they are related to have created such an unsafe situation for the world. Their evaluation is marsupial, not the normal the world has seen and has. Is it so difficult to get live weather feeds from around globe onto a plane? Even if the plane had everything working 100% as it always is, does it stand any change taking a Cb head on if it hits a blind spot as you indicated? NO

    Thanks for the post. Have added you to my regular Google search list. Regards.

  16. everglades Says:

    Mr. O’Brien –
    You continue to bring a knowledge filled reporting to this tragedy and I thank you for it.
    My father was killed over the Everglades 46 years ago in a strikingly similar accident and I think your reporting would be informed by reading over the CAB report and other information I’ve collected over the years.

    What’s the best way to forward it to you?

  17. everglades Says:

    So you’re not interested in the information about the everglades crash?

  18. everglades Says:

    I’d appreciate a response at Thx

  19. stewart Says:

    Any chance that they had XM weather(or something like it)on board?
    I assume with that they could have seen sat images. Had they elected a lower altitude to increase the speed margin would they still have had enough fuel to make the trip?

    Great story and great supporting data.

  20. mysguyded Says:

    Great article. I’ve worked in the aviation industry “Avionics” for many years, both military and commercial. Most people may be unaware of the fact that the pitot tubes are heated and should never ice up, they get screaming hot to the touch at sea level.

  21. mzen Says:

    Good article Miles,

    Couple of things here: the service ceiling for that transport is about 40k feet. At that altitude the sound barrier is around 660 mph and the max cruse speed for that plane is about 550 mph, so they were pretty well within the lift/drag envelope under normal conditions. The mission profile for the flight would be parabolic, i.e. rising to an elevation close to the service ceiling, burning fuel weight, and then descending. That profile is standard for nearly all commercial flights. As you said, the reason they do this is to conserve fuel costs.

    As a researcher of accident reports, most often it is easiest to blame the tragedy on human error (see Alpha Piper, Therac-25, TMI, etc.) but this is not always the case. If you study accident reports (which, Miles, I am sure you do) many accidents come down to a culture of indiscretion, meaning that the operators were probably doing their jobs, however the manner in which the job is to be carried out is content only with the bottom line. In this case that is fuel savings.

    Another possibility, which I believe another post mentioned, is that the true airspeed of the transport may have been misread or the feedback was false.

    In closing, I think the pilots were probably correctly carrying out their designed mission. By the time they realized the extent of the severity of the storm it was most likely too late. A very unfortunate event in any case. What a tragedy. Thank you for the article.


  22. Rachel King Says:

    I have to agree that was chilling and opened my eyes to the extremities of trans-Atlantic travel. But in any case, it it’s good to know…even if there’s nothing I can really do when I’m in a plane.

  23. billyrawl Says:

    Thanks for your post Mr. O’Brien, which shows a good grasp of aircraft technology and airline operations. Things have obviously changed since I started my career as a pilot for a U.S. international airline about 44 years ago at the beginning of the jet age. I’ve long since retired, of course, but a couple of possibilities come to mind.

    When I was flying I thought nothing of deviating hundreds of miles to avoid weather not nearly as severe as that experienced by this crew. As a former colleague reminded me recently, on a site that several former pilots from our now-defunct airline frequent, we were paid by the hour, and not paid extra to penetrate dangerous weather. Perhaps this crew felt that they needed to take this risk because they were not provided with adequate fuel to deviate the distance necessary to safely traverse this weather, but if that were the case, the time-honored “180” and landing at the nearest suitable airport is always an option, provided they had the fuel available to do that, which I assume they had. Of course, the 180 must be accomplished BEFORE entering the weather for reasons you stated. Then again, should they have done so, I wonder how their company management would have reacted.

    Another possibility I’ve thought about is that the lightning may have disabled the “fly-by-wire” boxes, and since they had no mechanical connections between their controls and the control surfaces, they were screwed. All the above is speculation, of course, and let’s hope they recover the recorders so that they’ll learn what happened here.

    Re: Fuel. When I started my career the Captain had the authority to increase his fuel load as much as he felt was necessary, no questions asked. Later on, after the fuel shocks of the ’70’s, we started getting arguments from dispatchers since fuel = weight = money. For my part, I won all those arguments, since I refused to take any flight without the fuel on board that in my judgment (NOT that of a dispatcher safely on the ground) was necessary to operate a safe flight. I always studied the weather charts carefully beforehand, and never ordered extra fuel without a good reason, but once I did no one was going to talk me out of it, and many ground-bound dispatchers tried. There were many occasions I was glad to have the extra fuel, as well as many whenI didn’t need it, but the consequences of having less that I needed were obviously much less severe than having more than I needed (Old pilot saying: “The only time you have too much fuel is when you’re on fire.”)

    Finally, one of your former colleagues, Mr. Charles Bierbauer, is a professor at our local university. I encounter him occasionally at social occasions.

    • Miles O'Brien Says:

      Very good points. Of course even with a lot of fuel a deviation would have likely forced an intermediate stop since the convective line was so large. A 180 was the better part of valor in this case. Give Charles my regards.

  24. fredrikk Says:

    Hi Miles,

    I do not understand your small speed window of 25 knots. I believe the window is larger than this for an A-330 cruising at say 35000 ft. (Also, the flight envelope shown must be for another aircraft and shows altitudes above 50000 ft).
    Otherwise, thanks for your article.

  25. brianj Says:

    Hey Miles, just wanted to say I was a fan of yours from CNN (as a former Space Camp nerd) and I’m glad you’ve found a home here. I’ll be checking in daily.

  26. testflyjets Says:


    Thanks for bringing a lot of information together on an otherwise complicated topic. Understanding what caused this tragedy is, obviously, one of the keys to preventing similar accidents in the future.

    I need to correct an inaccuracy regarding the “coffin corner” concept for an airliner. Just curious but what was your data source for the statement “AF 447 was flying through the eye of a speed needle only about 25 knots (28 mph) wide”?

    The range between stall speed and Mach buffet for an A330 (and all similar commercial jets) is much, much larger than 25 knots.

    The A330’s indicated airspeed would have been around 280 knots or so for a Mach 0.8 cruising speed at 35,000 feet. Stall speed would probably have been around 130 knots, a difference of 150 knots.

    Stall speed is purely a function of angle of attack and is given in terms of indicated airspeed, not true airspeed — the wing doesn’t know its at 35,000 feet so long as the same number of air molecules are passing over it, generating lift.

    It’s true that 120 knots Indicated Airspeed equals a much higher True Airspeed at 35,000′ than at sea level, but it’s irrelevant.

    The maximum Mach number (Mmo) for an A330 is around 0.86 M, and very bad things due to Mach buffet (like the tail coming off) wouldn’t happen until well above Mach 0.90 or so, another 40 or so knots higher.

    I find it difficult to call a nearly 200 knot spread between Mach overspeed and aerodynamic stall a coffin corner. Overstating the case doesn’t do much to ease the fears of “white knuckle flyers”.

    This is quite ironic because of the “flight envelope” graphic you included in the article, which you commented was just an example. It sure looks like one for the early U-2 (the airplane Frank Powers flew over Russia), and if any jet on the planet has a coffin corner, it’s the Deuce (former U-2 driver here). But a Deuce and an A330 are different animals indeed and an A330 isn’t flown “through the eye of a needle” – that would be bad for business.

    Also, no commercial jet operated within its published limits, even at its service ceiling, would stall simply trying to make a turn. Stall speed does go up in a turn due to the higher angle of attack required to maintain level flight. They might have to limit their bank angle to 20 degrees or so, but they could certainly turn.

    The real problem with a large jet like an A330 is what happens when you find yourself upside down at night in extreme turbulence because you penetrated a thunderstorm.

    If the pilots found themselves in this situation while simultaneously battling electrical failures that disabled their primary flight instruments or pitot-static system failures that caused erroneous airspeed indications it is entirely conceivable that they got into an unusual attitude from which they could not recover without ripping the wings off or snapping the jet in two due to an excessive over-g.

    I completely agree with you on the issue of pilots not being familiar enough with how to operate the aircraft’s weather radar. Even in this day of “high technology made simple” most weather radars still require a fair amount of experience and finesse to really use them well (as I know from crossing the Atlantic in the venerable C-141B). Some pilots simply don’t encounter much weather on which to practice.

    Regarding the radar, there’s usually plenty of lightning in and near a developing thunderstorm. What’s a very common spot on a large jet to get hit by a lightning strike? The radome, the non-metallic cover on the nose of the jet that sits in front of the weather radar dish.

    A strike to the dish would probably disable the weather radar, so the pilots would be flying blind from the perspective of being able to “see” the weather ahead of them, further complicating their attempts to avoid storm cells.

    Furthermore, if they were trying to shoot the gap between two cells using the weather radar, got into trouble and then tried to turn around to get out of the weather they could easily have found that their turn radius was large enough (probably on the order of 12-15 miles for a 180 degree turn) to put them right into one of the cells they were trying to avoid.

    Finally, one of the other comments here mentioned not being able to get “approval” for a diversion from the flight-planned route due to the use of HF radios and the non-radar oceanic environment.

    Thunderstorm avoidance can rapidly escalate into an emergency procedure and does not require an ATC clearance.

    No sane pilot would knowingly fly into a thunderstorm waiting for an oceanic controller to “approve” their deviation. You’d deviate, maybe change altitude by 500 feet just to be extra safe and then rely on TCAS not to hit somebody else doing the same thing coming the other way.

    Telling ATC what you did after the fact is a much lower priority than avoiding one of nature’s most significant weather phenomena.

    In general, flying on a commercial jet is about one of the safest things you can do. Folks just need to remember that, even when tragedies like this one occur.

    Thanks again for the report.

    – Chris

    • Miles O'Brien Says:

      Thanks Chris – I will check my numbers – and take down that graph.

    • rodgergriffith Says:

      Initial buffet boundary charts for large commercial aircraft show a much narrower margin than testflyjets suggests. For a 727 at 350 in zero turbulance it would be from 212K to 290K as the weight,turbulance or bank angle changes these numbers converge. With weight still high in initial cruise, severe turbulance and possible unusual attitude the convergence could well have put them in an unsustainable flight situation. Lack of radar or reliable speed information as well as violent buffeting would, if present, contribute to an impossible flight regime. If the verticle stabilizer was compromised all bets were off.
      As a retired AA Capt. I would speculate that all efforts will be made to cover up the inherent flaws of the A/C and once again blame the pilot. They are right in that regard only in the issue of AVOIDANCE. BE the CAPTAIN!

  27. glenn38 Says:

    I’ve always found the Nov 12, 2001 A300 rudder failure (flight 587?) explanation to be less than satisfying. If Airbus did such good job limiting FBW control surface travel, why did a low and slow flight in very predictable jet wake lead to a total vertical stabilizer / rudder loss? It just seems totally unacceptable, yet it was accepted and the co-pilot, now dead and mute, was blamed. I’m wary of more of the same with 447. I’m afraid that Airbus and Air France will gladly welcome frozen pitot tubes supposedly leading to relaxed rudder travel restrictions. It’s easy to fix. The alternate and very real possibility of early composite material component failure with an as yet undetermined failure mode means grounding most of the Airbus fleet until it’s resolved. Boeing would also be affected. All of us in commercial endeavors who depend on machine reliability hope to have easy resolutions to our problems, but reality and experience teach that what is not properly resolved today, will return and often with vengence. It’s extremely important to understand the sequence of failures in flight 447. The recorders must be found. In the meantime, we cannot accept a glossy review of the potential causes. Thanks for the well reasoned and fact based articles. Every other article in the media has been mere recital of the company line, and not true reporting.

  28. drewgolden Says:


    I always enjoyed your commentary on that Cable News Network and always felt you were somehow being held back.

    This article proves to me that you were.

    Plain, simple explaination that we can all understand – and yet very technical and cool in a geeky kind of way.

    I appreciate what you are doing and encourage you to keep it up. I am convinced that the AF 447 either ripped apart or stalled.

    30mph is a very small margin of error!!!!!

    Thanks for the excellent reporting!


  29. fredrikk Says:


    I strongly suggest that you make corrections directly in your coffin corner story (not only via author’s comments) to provide proper info for the speed window for A330 at 35000 ft (see comments by testflyjets and myself). You should also directly in your story point out that the flight envelope graph is for a completely different aircraft, and is for illustration only. Otherwise plenty of new readers will be misinformed.

  30. dougb Says:

    “Coffin corner” sounds great for a dramatic story, but I don’t see how it applies here. The term refers to the altitude at which a plane’s stall speed is the same as it’s critical mach number(exceeding which results in loss of control). For the A330, this altitude would be over 65,000 feet, or more than 23,000 feet higher than the plane is capable of flying. At 35,000 feet, there is a more than 100 KIAS (knots indicated airspeed) difference between the plane’s clean stall speed and it’s never exceed speed, not a mere 28 mph. At cruising speed (Mach 0.82) the plane may be flying within a few tens of knots of it’s never exceed speed (Mach 0.90, I believe), and an erroneous airspeed indication might cause the pilot or autopilot to push the nose down and gain too much speed, but that’s got nothing to do with the “coffin corner”. Throwing terms like that around is going to do nothing but scare your readers for no good reason. Unless something goes horribly wrong, transport aircraft fly well within their operating parameters at all times. It’s a slight different story with small business jets that may operate at higher altitudes, closer to their “coffin corner” (but not dangerously close).

    • Miles O'Brien Says:

      Understand. And if you read my piece I make that point.

    • rodgergriffith Says:

      You are wrong about the margins. See and its lesson on Buffet Boundary charts. Those charts are the bible for high altitude flight at high gross weight in turbulance with the possible need for maneuvering. Yes Virginia ther is a coffin corner!

  31. vlemos Says:

    Your rationale makes sense, if you were looking at a plane crash over the US or Europe. But you have to put things in perspective and realize that ground and flight operations are not the same in other parts of the world. Including Brazil. The clues to what happen are probably on the ground: the cargo manifest (AirWay bill) and aircraft maintenance records. Why has AirFrance not made this information public? The carrier is ultimaly responsible for ensuring passenger safety, not Airbus I’m affraid. Yesterday, the Brazilian newspaper “O Globo” reported that there were over 1,000 lbs of papaya in the aircraft, together with similar quantities of other fruits accommodated in woden pallets. These pallates were the first pieces of wreckage spoted in the Ocean. They were probably inside Aluminium containers, and it is common to have export goods. But What happens to the aircraft structure if there was mass imbalance during such severe weather conditions? What other goods were aboard? Has the possibility of a fire in the cargo bay been considered? Recall ValueJet (AirTrans) crash in the Everglades in the mid-90’s. NTSB did find improper handling of canisters was linked to a fire. Sadly enough, 228 people died but a lot of attention has been paid to the technical aspects – as you say, a “paradox of simplicity”.

    But no attention has been paid to AirFrance’s operational procedures and flight safety. This aircraft is flow from Paris to Rio the day before as flight number AF444 and then returned to Paris the next day as AF447. Do you honestly believe there is infrastructure in Rio to fix an Airbus A330-200? Wouldn’t you think it is more likely the aircraft may have been flying with a few problems — not considered to be critical, but may have become single point failure during the turbulence? Some “lateral” thinking is needed so that people look into some unsual scenarios. Presumably, this was not the first time the pitot sensors froze. What’s new there? There are a number of sub-systems in the aircraft and as you say the crew would have a handle on the situation. Structural damage strikes me as no-go. I work in the Space industry and can tell you the margins of safety employed in the design of aerospace structures is beyond belief. Unlike Space, aircraft are flight tested etc. If it wasn’t like that there would have been similar accidents in the past. But none have happened. You rightfully pointed out there were at least 12 other flights that night through the same route. Of course the weather patterns can change dramatically and there could be a unique situation.

    Have a look at this, maybe you can help. I lost a dear friend in that flight.

    • patricksmith Says:

      “But no attention has been paid to AirFrance’s operational procedures and flight safety. This aircraft is flow from Paris to Rio the day before as flight number AF444 and then returned to Paris the next day as AF447. Do you honestly believe there is infrastructure in Rio to fix an Airbus A330-200?”

      Well, I suppose it depends on the nature of the repair, exactly, but why not? Brazilian carriers operate and maintain all the various Airbus and Boeing models, and presumably could have been available to perform some contract mx work. This sort of thing happens all the time, all around the globe. (And remember, Brazil has a very long and proud history of aviation. Heck, the country builds more high-tech regional jets than anybody in the world.) There is no reason to think a carrier or mx station at Rio would not be equipped to handle an A330 issue.

      Patrick Smith

  32. patricksmith Says:

    Miles used the same graphics I did in illustrating my most recent column at Tim Vasquez ought to be getting royalties.

    “Coffin Corner” for dummies…

    Flying at 35,000 feet, airspeed is critical to avoid stalls. A plane stalls when, roughly put, the wing runs out of lift. But not only are there low speed stalls, there are high-speed “shock stalls” as well. It’s not only a question of how efficiently the wing is moving through the air, but how fast the air itself is flowing * around the wing. * As this flow nears the speed of sound, a shock wave builds, disrupting the airflow and destroying lift. A stall.

    Thus, at upper altitudes, where the air is very thin, a plane can find itself caught in an aerodynamic paradox: the higher it flies, the faster it needs to go to maintain lift; but the faster it goes, the closer it gets to that shock wave. You’re stuck between going too fast and too slow at (almost) the same time. This realm is called “coffin corner” — a scary buzzword that has been making the media rounds of late.

    Crews will calculate buffer speeds to keep them at a safe distance from this proverbial razor’s edge, and it’s only relevant at the limits of the performance envelope, which is not where airliners normally hang around.

    For flight 447, however, things were presumably anything but normal. A rapid increase in engine power, owed to a faulty pitot input, could have caused a high-speed stall within seconds. And because of the weird aerodynamics of these stalls, recovery can be exceptionally difficult – sometimes impossible – especially if dealing with heavy turbulence, electrical faults, and flight control failures at the same time. They would have had little chance.

    Just one gripe though…

    Writes Miles: “So while you are napping, eating or watching a movie on that flight to LAX, you should know the plane you are flying is cruising along at the ratty edge of its capabilities. Why? Money. The higher an airliner flies, the better gas mileage it gets.”

    Well, like I said, this is not where planes normally hang around. And actually, “economy” is probably a better word than “money” in this context. Jets are designed to fly this way, and have doing so safety for the last 50 years.

    Patrick Smith

  33. jrgiguere Says:

    “Fortes turbulences” means “strong”

  34. Essay of the Week: 6/14/09-6/20/09 » The Ethereal Voice Says:

    […] Miles O’Brien will educate you on lot of things that most fliers prefer not to think about. Part one is here. Part two is […]

  35. libtree09 Says:

    Chilling but as someone who used to travel quite a bit I’m a bit confused or misinformed. I was not aware that a storm could be found at forty thousand feet. I recall flying over many terrible storms and even taking off with white knuckles in one of the worse in Australian history. Nor did I know about the ratty edges of flight although flying in a aluminum bus whose structure could no doubt be pierced by a good screwdriver is not very reassuring either. Lastly using radar for weather does sound a bit 1940’s.

  36. PhotoGeek Week#93 The Goose that flies backwards. | Photogeek. Podcast for the geek photographer, professional, prosumer or keen beginner Says:

    […] Air France crash theory. […]

  37. mcowan Says:

    The recent photo of the suspended Flight 447 ver. stab. and another photo showing an intact rudder is evidence that the bending moment at the time of the break was entirely along the x-axis. This is consistent with a postulated excessive lateral aerodynamic overload and a large yaw angle relative to the airstream.

    Ice in/on the pitot tubes is not likely. The PTs have powerful electric heaters, and the q at 35K, .81 M is approx 3.5psi with zero airflow inside the tube. The most likely candidate for the anomalous air data signals is the system design. Triplex (not recommended but used in the Airbus) redundant digital FBW control systems are fail op/fail op based on self test/fail. The sensor preprocessing is: A/D sample and digitize in each channel – cross channel differences formed and conditioned in each channel – differences are voted upon – odd channel ignored – all channels agree on best choice. The result is that all 3 channels give identical commands to their individual hydraulic servovalves.

    POSTULATED CAUSE of 447, 587 crashes and logged anomalous incidents. The yaw damper (see same) and 1st bodybending mode feedback loop undergoe sustained oscillations of increasing amplitude to the point that aerodynamic loads on the plane exceed ultimate strength of the structure.

    HYPOTHETICAL SCENARIO: 447 enters region of severe turbulence. Intermittent YD/BBM1 oscillations develop and produce large excursions in pitch and yaw (the YD couples roll and yaw – see Dutch Roll) angles which cause pitot outputs to disagree. (They don’t see the same airstream) Lateral loads are placed on the Ver.stab. One channel of air data sensor is voted out. Over the next few minutes a second channel is voted out. The remaining channel is degraded and the poor data doesn’t help the situation. Fatigue and or a sudden big side load breaks off the ver. stab. Aircraft goes broadside to the airstream and begins to break up.

  38. Notizie e curiosità in pillole « Paoblog’s Weblog Says:

    […] che aspetto ha un articolo d’aviazione divulgativo scritto come si deve, leggete cosa scrive Miles O’Brien a proposito del Volo 447 precipitato nell’Atlantico per ragioni ancora non […]

  39. golf3400 Says:

    Good analasys but it true for conventionnal aircraft but quite different for fly by wire A330. If the aerodynamic is the same, the controls are working by a complete different way, especialy when the plane is in a flight position outside the normal flight envelop.

  40. Un avi Says:

    […] se aclaran ahora: Lean, co

  41. True/Slant Anniversary: Some of the best analysis and perspective published in our first year - Michael Roston - Newsbroke - True/Slant Says:

    […] aviation to write a pair of brilliant, in-depth blog posts – the first on June 1, the second a week later – that debunked a lot of the myths and explained a lot of the facts about Airbus’s […]

  42. PhotoGeek Week#93 The Goose that flies backwards. | Says:

    […] Air France crash theory. […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: