I’m wondering, given Airbus history, if it will turn out to be a problem with the fly-by-wire flight-control system.
17 thoughts on “The Airbus Crash In Europe”
Comments are closed.
I’m wondering, given Airbus history, if it will turn out to be a problem with the fly-by-wire flight-control system.
Comments are closed.
There are literally dozens (if not more) possible causes for an airliner crash. At this moment, all we know is that the plane crashed and it’s unlikely anyone survived. The Smithsonian Channel has a series called “Air Disasters” that is about accident investigations. As a pilot, I watch the show often. The moral of almost every episode is that it’s almost never what people first thought or reported. As hard as it is, we’ll likely have to wait a long time to learn what caused that plane to crash. They can sometimes make an initial determination in a short time but most of the time, it takes months or even years to determine a cause. Sometimes, they never know. At least this crash happened over land. It’s in very rough, snow-covered country so the investigation will be difficult but not as bad as one that’s under water.
Isn’t another moral of most disasters that a cascade and/or confluence of small mistakes and events usually end up combining to create a disaster, and that taking out just one or two of those things would have kept the disaster from happening?
For example, a softer initial landing in the leg from Bastion to Bagram, a more experienced crew, two extra straps per MRAP, etc. may all have kept something like National Air 102 from happening?
I wouldn’t be surprised if there were a number of small things that cascaded into one big problem for this crash, too. As you mentioned, the fact that it’s on land will make recovery easier than a water crash, but still not very easy.
Yes, the chain of events leading to an accident is the subject of a great deal of aviation safety training. Most pilots can handle a single problem. Professional airline pilots are trained to deal with very difficult circumstances (multiple mechanical/electronic problems along with bad weather, etc.). However, a chain of events and decisions, many of them seemingly inconsequential at the time, can result in an unrecoverable situation.
I’ve heard they’ve already recovered at least one of the orange “black boxes”. That’s good news. Unless there was something like an electrical power problem that knocked off power to the boxes*, investigators will soon have a wealth of information to sort through.
*There have been several accidents where power is lost suddenly and the boxes quit recording. These are normally catastrophic explosions like TWA 800, which seems not to be the case for this accident. For some reason, they don’t include backup batteries in most cockpit voice recorders and flight data recorders. I don’t know why.
I’ll also point out that oft times it’s not one problem, but rather a chain of events that’s to blame. Birgenair Flight 301 is a perfect example of that.
Initial reports that the plane descended for 8 minutes after reaching cruise, with no distress call from the cockpit, lead me to two theories: 1) the Payne Stewart scenario, where the crew was disabled by a failed pressure / temp life support, or 2) a deliberate human agent in the cockpit (how’s that for euphemism?)
You’re not kidding about the chain-of-events business. Way back in the ’70s, there was a mid-air on final approach to Moffett NAS, between a Navy P-3 and a NASA CV-990. The crash site was the 12th green of the Sunnyvale golf course, less than 1K from the runway.
The final report showed that a chain of at least 6 discrete errors over the preceding 45 minutes, both in the air and on the ground, led to the accident (which was, basically, both aircraft attempting to land on the same runway. Moffett has parallel runways.) One error, for example, involved a light twin taxiing the wrong way, more than 30 minutes before the crash.
About 2 decades ago, I used to frequent AirDisaster.com as a commenter. Then, I was very curious about system failures and ways to prevent them. Whenever a new event occurred, we speculated.
On this, I’m not interested in speculating. I have ideas of what it could be, but already a black box has been discovered. It won’t take long before we have much more information than just a radar track and a debris field.
Here’s what we knew, basically immediately, from the FlightRadar24 web site:
– it initially climbed to 30,000 ft, levelled briefly, then climbed to 38,000 ft, then after about five minutes started a steady and normal descent until it hit the ground.
– the flight track precisely matched that of the same flight the day before, INCLUDING at least one programmed heading change near Toulon/Marseille while it was descending.
– ground speed decreased from near 500 knots to just under 400 knots as it descended from 38,000 ft to 7000 ft.
Here’s deduction and speculation:
– it remained on autopilot. The pilots not only did not communicate with air traffic control, they did not make any attempt to divert to an airport. This suggests the pilots were either unaware of the descent, or incapacitated.
– the decreasing ground speed as it descended is expected, but it doesn’t look like enough for a normal descent. Normally descents are flown at constant Mach number until the IAS increases to 250 knots (from maybe 130 or so at cruise altitude), and then at a constant 250 knots IAS. The absolute maximum never exceed IAS is 350. Even with a tail wind and still at 7000 ft they must have been very close to that. But perhaps didn’t exceed it. They definitely blew well past 250 knots IAS.
– the climb to 38000 ft seems unusual for an A320, especially for what is only an 1100 km flight. On the other hand, 35000 or 36000 would not have surprised me. I don’t know what the cruise altitude was on that flight the previous day.
Pure guess: they suffered de-pressurisation, the crew set a higher than normal descent rate into the autopilot, and then became incapacitated when the supplementary oxygen system failed.
Some stuff seems off with your numbers: IAS at FL380 and Mach 0.8 should be in the vicinity of 280 KCAS (it’s CAS at those Machs, not IAS, and I’m doing that calc in my head); and tailwind doesn’t matter for airspeed. But the ground speeds you cite look reasonable at altitude and for a descent on full autopilot with Mach hold engaged, then hitting the yellow line (400 KGS is really fast for 7000 ft unless that’s AGL over high terrain instead of baro alt). A very odd descent to be sure.
To return to Rand’s original comments, Airbus has significant issues with their fly-by-wire design, especially in their management of redundancy and in their autopilot moding. Whether any of that contributed remains to be seen.
Ah, yes you’re correct on the speed at cruise altitude. There is less than a quarter of normal air pressure at 38000 feet, but I forgot the IAS goes as the square root of the pressure, not directly proportional to it.
Tail wind does not matter for airspeed, of course, but the figures broadcast by the ADS-B and received by FlightRadar24 are GPS ground speed, not airspeed.
All that we “know” “for certain” is that it wasn’t terrorism. That was announced almost before the plane was known to have crashed. In fact, you can probably just make a universal announcement, right now, that “no human death anywhere on earth at any time was caused by terrorists.” And you’d be almost right.
It will take a very long time to figure this one out, at least to the point where the public is informed. I’ve been involved in such investigations, and know why the process is so slow. It’s for good reason. When you know, though, it will be the best knowledge you can have.
That is, aside from TWA Flight 800…
So we now know the Captain was pounding on the locked cockpit door yelling, “Achtung, Comarade! Lass mich herrein!” (Yo, bro, let me in!)
And the American Airlines has had a rule where if someone on the flight deck has to use the lav, a flight attendant has stand guard at the door from inside the flight deck so there are two crew members on station at all times?
It seems people have at least talked about this kind of scenario?
While I think I notice Mfk’s tongue in cheek; I do find it interesting that people are already associating the co-pilot to some nefarious plot. I say this because early on, I saw many odd arguments for decompression and pilot incompacitation. Ignoring for a moment that airbus cockpit door lock may have 3 settings: unlock, code (that a crewmember can type to unlock, and locked (only unlocked from within cockpit). The pilot could have left the cockpit and the door locked behind him and then something incompacitated the co-pilot. No suicide, no homicide, no terrorism; just series of unfortunate events.
I thought the Flight Data Recorder storage had yet been found. I think you would need that data to confirm the co-pilot did something nefarious.
Reportedly, it was supposed to open with a code, but the co-pilot was overriding. It seems quite certain that it was deliberate.
The FDR is still missing, but according to Flightradar24, analysis of the ADS-B/ModeS data indicates the copilot set the autopilot to 100 feet after the pilot left the cockpit, and Fox is saying the CVR shows regular breathing from the copilot up until impact.
Setting the autopilot to 100 ft (actually the FR24 data says 96 ft) does seem pretty conclusive.
How the passengers could fail to be alarmed by the pilot desperately trying to open the door, I don’t know.
There is one known case, ever, of hijackers deliberately crashing a plane. (ok, several cases on the same day). And that can never happen again, for reasons many of us — and indeed the passengers of United 93 — have understood since 2001.
There are half a dozen cases in the last 15 or 20 years of pilots doing it. Quite probably including MH370.
We’d be better off banning cockpit doors entirely than turning them into fortresses.
Yeah the whole idea was more trouble than its worth. More people have died because of the barred doors than otherwise.