| BlueSky Business Aviation News | ||||||||||||||||||||||||||||||||
| . | ||||||||||||||||||||||||||||||||
Let me explain a bit. Winds are a part of what is called "atmospheric circulation" (the other parts are updrafts and downdrafts) and atmospheric circulation is driven by solar heating. I.e., the hotter it gets, the more circulation we can expect. Ergo, we get a lot more circulation near the equator than we do at the poles. Ever see a thunderstorm over the north pole? Well, there you go. Actually, thunderstorms do occur near the poles, but in comparison with the ITZ, polar storms are very rare. Now, the Hadley Cell is formed by the warm and very moist air near the equator. As that mass of air rises up, of course, it cools; the condensing moisture creating heavy rainfall. The warmer the air, the more moisture it can hold, and, therefore, the heavier the rainfall. As the water condenses and falls, each drop gives up what is called the heat of latency (sometimes giving up so much heat that the droplets are actually well below freezing in temperature, or "supercooled", and don't have enough internal energy to crystallize into ice).This released heat causes the now slightly warmer and drier air to rise farther and faster before it spreads out at the top of the cell, moving north or south into cooler, drier regions. As the air cools, it becomes denser and falls downward, creating a downdraft. This downdraft is the other edge of the Hadley Cell. It occurs in what is called the Horse Latitudes (long story having to do with the way seamen were paid). Because the air is flowing strongly upwards near the equator and downward near the Horse Latitudes, a strong, regular surface flow toward the equator is created. We call that flow "the trade winds." Thunderstorms are bad news Pilots have long known that thunderstorms can be deadly. The powerful, chaotic flows inside a thunderstorm can quite literally tear an airplane apart, even a modern one, and I'm not talking about surface micro-bursts here, but the fundamental flows within a thunderstorm have torn apart several modern commercial jets, including a 707 (which is a tank in comparison to most aircraft). And it isn't just the wind shear, either; the rapidly condensing moisture can easily form ice on wings, engine inlets, control surfaces, and pitot tubes. Remember, those tiny just-condensed droplets of water are below freezing in temperature, but aren't actually ice at this point. Splatting them across a hard surface like a wing leading edge provides more than enough energy to turn these super-cold droplets almost instantaneously into rime ice, which is difficult to shed. Pitot heat It is this tendency for ice to form at altitude that pitot tubes are heated in flight. Always. As the aircraft flies through the air, the pitot tube rams into those supercooled droplets, causing them to turn into ice. Sometimes the ice forms on the tip, sometimes it forms inside. The pitot heater is supposed to provide enough heat energy to liquify the newly formed ice, allowing it to run off (and out) as liquid water. A frozen static port causes the airplane to think it is flying steadily at a constant altitude. A frozen pitot port causes the airplane to think it is flying at a constant airspeed--until it either gains or loses a fair amount of altitude. Pitot what?
The holes that measure the static pressure and the hole that opens into the pitot tube proper are pretty small. If they are plugged by ice (or dirt or insects), then the pressure measurements are wrong, and the pilot's instruments read incorrectly. Worse, in today's modern computer controlled aircraft, if the computer can't get accurate data, it may well do something unexpected, this is particularly true (and hazardous) in aircraft controlled by computers with "envelope protection." Remember the US B-2 bomber that crashed in Guam because of water in its pitot system? More to the point, remember the crash of Birgenair Flight 301 due to a blocked pitot tube? It is very difficult for the pilot to recover the aircraft after a loss of the pitot-static system; in some cases, it may be impossible. Coupling Closely coupled computer systems have a number of different computers that talk to one another, passing data back and forth and often using the same data for different purposes such as navigation and flight direction. A fault in one module can propagate through the entire system rather like a digital ripple, causing other modules to make bad decisions or pass on flaky data to downstream modules, causing further disruption to other systems. For example, if the Air Data Inertial Reference Unit (ADIRU, we call it) gets bad data from frozen pitot tubes and passes that data on to the Electronic Flight Control System which will react as though that data is correct (because it can't tell the difference--see Hume), in some cases over-riding the pilots' inputs (and we won't know if this happened on Flight 447 until the FDR can be read). I've simplified things a bit, and there are a lot of technical caveats to that simple example, but it gives you the general idea. Tightly integrated, closely-coupled systems are the norm in today's heavy jet transport for a number of reasons including weight and performance. There has been an enormous amount of work done on making these systems as smart as possible to minimize the risks associated with close coupling, but things aren't yet perfect, obviously. Sounds crazy, doesn't it? Still with me? Let us integrate all that. A highly experience and competent crew were flying their modern computer-controlled aircraft into an area of powerful storms which create supercooled droplets of water, which turn into ice immediately after hitting some part of the aircraft. If there is enough ice, it blocks the pitot tube and or the static ports, causing the aircraft computers to do weird things. Still, it isn't quite as bad as it sounds. Commercial aircraft do this sort of thing every single day with an amazing level of safety. The crews use their on-board weather radar to avoid the worst of the storms by simply flying around them. One of the things we don't understand is why this crew apparently did not do that (we need the CVR to know) and exactly what happened as a result. Messages of failure The A330-200 that crashed had an ACARS on-board. The ACARS is a digital radio datalink from the aircraft's computers to their maintenance station. It has been in wide airline usage for some thirty years. The system can send messages automatically, without crew involvement. Those messages are usually related to flight phase or equipment malfunction. Flight 447's ACARS sent a four minute long string of messages documenting various system failures. Presumably, the string ended on impact. There were messages reporting the autopilot being disconnected (we need the FDR to understand why), malfunction of the pitot-static system, as well as damage to various parts of the airplane including the flight control computers, loss of electrical power, and loss of cabin pressurization--thought to be due to the airplane breaking up. The black boxes After each accident, the investigating agency--the National Transportation Safety Bureau (NTSB) here in the US, the Bureau d'Enquêtes et d'Analyses pour la Sécurité de l'Aviation Civile (BEA) in France--recovers the Flight Data Recorder (FDR) and the Cockpit Voice Recorder (CVR) boxes. Those two recorders are carried on-board jet transports, usually in the fuselage in the rear of the airplane, strictly to facilitate accident investigations. Oh, and they are actually orange, and not black, so they are easier to find. The CVR records about a half an hour of the sounds of the flight deck. It's not always the conversation between the crew that matters. The aural alarms can be heard on the CVR, as well as other sounds of interest to the investigators. The FDR records hours of detail on hundreds of different parameters, including switch position, control surface position, speed, altitude, engine power, and many more. Both are really needed in order to fully understand why the accident happened and how events unfolded. And those boxes could not be found, until this week, almost two full years and four searches after the crash. The courts It is exceedingly rare that a government would assemble four separate searches for a crashed airplane lost in the ocean. In this case, it was neither France nor Brazil paying for the search; it was Airbus and Air France. This struck me as very unusual, so I asked around a bit. There is a significant difference in the way the courts handle this sort of tragedy here in the US and the way it is handled in France. In the US, a civil investigation would be the usual result. In France, a criminal proceeding is the norm. In fact, manslaughter charges have been filed against both Airbus and Air France, which is perfectly proper in the French legal system. That may well be the motivation behind the search. Or it may be that until the crash has been fully analyzed, both companies are liable for the many pending lawsuits. In case you were wondering, France is responsible for the investigation because the aircraft was registered in France and the accident happened in international waters. Current theory From what I have read and from the people I correspond with, the best current theory is that the crew couldn't see the really intense storm on radar as it was masked behind a smaller, less intense storm that they could see. When the aircraft entered the larger storm, the supercooled water droplets rapidly iced over the pitot-static system, causing a wide variety of closely-coupled computer failures due to disagreement between data streams--this can be extremely difficult to sort out, even with the experienced crew Flight 447 had. You may recall the QANTAS A380 incident where the engine blew up in flight. That caused some 57 separate failure messages which took over an hour for their highly experienced crew to evaluate. In calm air. On a clear day. Without good airspeed and altitude data, the pilots flew into a high speed stall, according to this theory, from which they could not escape. This high speed stall led inevitably to the crash. Stalls Let me digress just a moment for the non-aviators. When your car stalls, the engine stops, and the car rolls to a stop. In aviation, "stall" generally has nothing to do with the engines. It means that the airflow over the wings which produces necessary lift has somehow been compromised. When an airplane stalls, it drops rapidly because the wings aren't producing the required lift, but the gravity is still working. Stalls can happen at any speed, really, though they are much more common at low speeds. Stalls can also happen if the wing gets iced over; this changes the shape of the airfoil, which changes its flight characteristics, and never in a good way. At high altitudes, the minimum speed the airplane must have in order to stay in the air (the "stall speed") is quite close to the maximum speed (exceeding the maximum speed can cause structural failure of the aircraft; it isn't merely a matter of regulation). It takes surprisingly little in those tight circumstances for the wing to stall, or lose lift, and if the crew have wrong or no information on the aircraft's actual speed on a dark night in heavy turbulence, then they probably will not be able to recover before it crashes. Unanswered questions Of course, there are other theories, but this one seems to match what little we actually know about the accident, and matches up well with what has been found so far during this fourth and last search. The debris field is unusually compact, indicating that the airplane was intact when it hit the water and that it hit in a normal flight attitude, implying a high speed stall. But, we need the information contained in the CVR and the FDR to truly understand what happened. Was the airplane in a flat spin? Was it really stalled? What systems had actually failed? Why did they fail? What did the crew actually know at the time? What actions did they take? How did the aircraft's systems help or hinder? There are many, many questions that need to be answered before we can say we understand what happened to Air France Flight 447. The pitot tube furore The pitot tubes, like almost every other piece of equipment on a heavy jet aircraft, are made by two different vendors. Thales and Goodrich both make pitot tubes for the A330. The same model of pitot tube made by Thales is installed on both the A320 and the A330. The A320 is a much smaller airliner. The A320 had had a series of malfunctions related to pitot icing involving early production Thales tubes and there was a program to replace them with later production tubes that were thought to solve the problem, though in another report, the later production tubes had not been redesigned to better resist in flight icing, but to address a different problem altogether. Flight tests showed that the Thales tubes were much more susceptible to icing than the Goodrich model. There are cries to "Just replace them all!" but that ignores the reality of parts availability. Airbus issued a service bulletin to replace the tubes when parts become available, but allowed continued operation with existing parts until then. The only other alternative is to ground the fleet, which is mind bogglingly expensive and not usually necessary. The key concept The thing to keep in mind is that an accident is rarely due to a single event or condition. Usually, there are three or four or five or more events or conditions that have to happen before an accident happens. Here we have bad weather, probably iced up pitot tubes, and a series of likely computer failures due in part to a probably iced over pitot tube (and due in part to the closely-coupled nature of the design). But, I only fly on a Gulfstream! Right. Any idea how many computers there are on a Gulfstream? Exactly. And newer aircraft will only have more computers, even more tightly coupled. Given the capabilities of current and soon to be available business jets, you could also easily find yourself in that same Intertropical Convergence Zone at night in bad weather. And there is nothing magical about the pitot tubes on a Gulfstream that makes them more resistant to icing than those on other aircraft. In other words, we in business aviation have a distinct interest in understanding the larger issues surrounding the crash of Air France Flight 447 because we may learn something that will save the lives of our passengers and crews. The next few months could be very interesting. Terry Drinkard is currently consulting on an aviation start-up. His interests and desire are being involved in cool developments around airplanes and in the aviation industry. Usually working as a contract heavy structures engineer, he has held positions with Boeing and Gulfstream Aerospace and has years of experience in the MRO world. Terry’s areas of specialty are aircraft design, development, manufacturing, maintenance, and modification; lean manufacturing; Six-sigma; worker-directed teams; project management; organization development and start-ups. Terry welcomes your comments, questions or feedback. You may contact him via terry.drinkard@blueskynews.aero Other recent articles by Terry Drinkard:
|
||||||||||||||||||||||||||||||||
