Air France passenger jet drops off radar

  • Thread starter Thread starter NM
  • Start date Start date
Status
Not open for further replies.
What I dont understand is how a pilot can be essentially stalling the aircraft and not realise. I have flown only single engine trainers so my only experience if forced stalls during training but surely - the nose is up and power is doing nothing so the nose needs to come down?
Pitch black ... very poor weather ... lots of turbulence ... incorrect instrument readings ... confusion about alarms/alerts ... poor understanding of "what's it doing now" ... holes align in the swiss cheese :(
 
One of the problems with the Airbus is their aim to make it so easy to fly, that your milkman could do it. In so doing, they've merrily removed many of the cues that you are used to, and replaced them with 'annunciations' which are akin to trees hidden in the forest.

Whilst thrust lever position isn't an indication of what the engines are doing, it's a pretty good clue to what the autothrust is commanding...but that clue is removed and replaced by non 'back driven' levers, that just sit there. Go beyond that a bit, and once people have been flying them for a while, they simply stop looking at the engine power settings, 'cos the aircraft will look after it. They've even made it difficult to manipulate the power if needed (for instance, slight adjustments in HOLD mode are common in a Boeing, and impossible in an Airbus).

ECAM works a treat for a single problem, or for well understood scenarios which have a controlled sequence of failures. It is hopeless and confusing when multiple things go wrong at once. It prioritises a couple of things well, but give it more than that and the outcome will not necessarily be right, or repeatable.

Control laws...the aircraft starts with an odd thought....'don't touch it'. That's literally what you do when manually flying..put it in the right spot, and LET GO OF THE STICK. The problem is that it is easy, even with everything working, to end up overcontrolling. The joystick is extremely light, but has a very slight delay in any response. In normal law you can't stall it (!). Reversions to other laws aren't necessarily as obvious as you might expect. ECAM messages relating to them may well be hidden in amongst that forest. And the indications on the PFD are pretty subtle. Add to that that the aircraft flies totally differently in these other laws, that you have no aileron trim (and may need it), and a reversion isn't necessarily something to take lightly.
 
One of the problems with the Airbus is their aim to make it so easy to fly, that your milkman could do it. In so doing, they've merrily removed many of the cues that you are used to, and replaced them with 'annunciations' which are akin to trees hidden in the forest.

Whilst thrust lever position isn't an indication of what the engines are doing, it's a pretty good clue to what the autothrust is commanding...but that clue is removed and replaced by non 'back driven' levers, that just sit there. Go beyond that a bit, and once people have been flying them for a while, they simply stop looking at the engine power settings, 'cos the aircraft will look after it. They've even made it difficult to manipulate the power if needed (for instance, slight adjustments in HOLD mode are common in a Boeing, and impossible in an Airbus).

ECAM works a treat for a single problem, or for well understood scenarios which have a controlled sequence of failures. It is hopeless and confusing when multiple things go wrong at once. It prioritises a couple of things well, but give it more than that and the outcome will not necessarily be right, or repeatable.

Control laws...the aircraft starts with an odd thought....'don't touch it'. That's literally what you do when manually flying..put it in the right spot, and LET GO OF THE STICK. The problem is that it is easy, even with everything working, to end up overcontrolling. The joystick is extremely light, but has a very slight delay in any response. In normal law you can't stall it (!). Reversions to other laws aren't necessarily as obvious as you might expect. ECAM messages relating to them may well be hidden in amongst that forest. And the indications on the PFD are pretty subtle. Add to that that the aircraft flies totally differently in these other laws, that you have no aileron trim (and may need it), and a reversion isn't necessarily something to take lightly.

jb747 you are a very well paid and knowledgeable milkman. :p
 
JB747 - I think we are very lucky to have such a knowledgable person posting on this forum. As a lay person whose only experience of flying is as a passenger I find your comments and commentary absolutely fascinating and food for thought, even if I only understand about half of what you are saying. Its really interesting to hear an opinion from the other side of the coughpit door. Keep up the good work, your input is really valued! :)

Cheers

Timmi
 
Has there been studies done on feasibility of parachutes on planes?

I do remember seeing b&w footage of a prototype system. Each seat would run on a track backwards and out the rear side door with the passenger strapped in and parachute built into the seat.

It worked, but it was in something between a DC3 and DC6.
 
One of the problems with the Airbus is their aim to make it so easy to fly, that your milkman could do it.


Wish I still had a milkman......just like I wish the pilots control stick was still actually connected to the mechanicals that power the plane. :shock:
 
......just like I wish the pilots control stick was still actually connected to the mechanicals that power the plane. :shock:

There's nothing wrong, per se, with fly by wire. It offers many advantages over traditional systems. But, Boeing and Airbus have taken a very different track to its implementation.

Traditional aircraft controls aren't quite as directly connected as you might imagine. Sure there are cables and pulleys, but they connect to hydraulic jacks, so the aircraft are actually 'fly by hydraulics'.
 
http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.en.pdf

This is the reference to some initial information being provided by the BEA. Whilst not a great deal is there, it makes for pretty horrifying reading.

Basically, as I see it, the aircraft pitot/static system has played up, and caused the aircraft to revert to alternate law. It has issued a confusing bunch of warnings, coincident with loss of autopilot, autothrust and flight directors.

I do not understand why the aircraft has climbed steeply at the start of the event, and can only surmise that the FO flying has caused that. A 7000 fpm climb is extremely steep. It would require an attitude change of about 5 plus degrees to achieve that...but that sort of climb rate is well beyond the aircraft's performance, so it will be accompanied by a rapid bleed off of speed. As the aircraft speed bled, in alternate law, the system would be autotrimming the tailplane. The trimmable tailplane has much more authority than the elevators. It's interesting that it stayed in a large nose up position until impact. (The tailplane was also involved in the ANZ A320 crash, although in that case the aircraft was in direct law, when manual trimming of the tailplane is required).

The constant nose up control inputs beggar belief. At altitude an aircraft is not going to recover from a stall with any nose high attitude...but 5 to 10 nose down would have had it accelerating rapidly out of the stall.

Because Airbus control sticks are not interlinked at all, the other pilot has no idea what inputs are being made. Yet another cue that has been intentionally removed.

Discussion of this with a couple of other pilots came up with a couple of thoughts. Firstly, until very recently, the memory item response to loss of airspeed had, as a first item, selection of 5 degrees nose up, and TOGA thrust, but it had no consideration of what the aircraft was actually doing, so the initial climb could well have been a 'perfect' response to their checklist. Secondly, in normal situations, the use of up to full backsick will have no adverse affects at all...you simply cannot stall the aircraft. All well and good, but it means you need totally distinct responses to flying the aircraft depending upon the law it's in...and you have to realise that it has changed.
 
Read our AFF credit card guides and start earning more points now.

AFF Supporters can remove this and all advertisements

There's nothing wrong, per se, with fly by wire. It offers many advantages over traditional systems. But, Boeing and Airbus have taken a very different track to its implementation.

Traditional aircraft controls aren't quite as directly connected as you might imagine. Sure there are cables and pulleys, but they connect to hydraulic jacks, so the aircraft are actually 'fly by hydraulics'.

I do realise.

Did you know, Merc Benz has a car they developed where the controls including the steering are not connected by any linkages, all by wire. Expensive to develop but cheaper to make because it's simply electric motors and wiring. But, no country will give it certification at present because of the risk of it failing, while I'd be happy to drive it I would prefer my steering wheel connected to the wheels. :cool: (I would think the real risk is not getting is serviced by M'Benz the cause of a failure being the maintenance itself)
 
http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.en.pdf

Discussion of this with a couple of other pilots came up with a couple of thoughts. Firstly, until very recently, the memory item response to loss of airspeed had, as a first item, selection of 5 degrees nose up, and TOGA thrust, but it had no consideration of what the aircraft was actually doing, so the initial climb could well have been a 'perfect' response to their checklist. Secondly, in normal situations, the use of up to full backsick will have no adverse affects at all...you simply cannot stall the aircraft. All well and good, but it means you need totally distinct responses to flying the aircraft depending upon the law it's in...and you have to realise that it has changed.

I find it odd that fully trained pilots would hold the nose up for so long willingly unless they thought it needed it. As the report said that one of the pilots called out a change in flight law, I would assume they know the difference.

Do you think there might be of been problems with the control-stick or system behind it?
 
Last edited by a moderator:
I do realise.

Did you know, Merc Benz has a car they developed where the controls including the steering are not connected by any linkages, all by wire. Expensive to develop but cheaper to make because it's simply electric motors and wiring. But, no country will give it certification at present because of the risk of it failing, while I'd be happy to drive it I would prefer my steering wheel connected to the wheels. :cool: (I would think the real risk is not getting is serviced by M'Benz the cause of a failure being the maintenance itself)

my car is drive by wire and it's great!

The air france crash looks similar to the aero peru 757 crash.........
 
I find it odd that fully trained pilots would hold the nose up for so long willingly unless they thought it needed it. As the report said that one of the pilots called out a change in flight law, I would assume they know the difference.
You aren't the only one who finds that odd. But, this quote, in reference to the person flying, and if it is accurate at all, might start to explain it.
'Baby' pilot at controls of doomed Air France Airbus | The Australian
“The European airlines select people with virtually no flight time at all and train them pretty much from the ground up,” he said.
“They are 'company babies' who rise up through the organisation. Whereas if you get your experience in the navy or air force, there's an emphasis on trial by fire.”
So, apparently, the aircraft was actually being flown, at least for much of the event, by the most junior of the pilots. More of less the equivalent of the 200 hour cadets that Jetstar thinks is such a good idea. He may have had more flight hours, but in all honesty, you do not learn how to fly in the coughpit of an airliner, nor in a simulator. You do learn how to manage systems, but, not how to fly.

As to why he was flying it, and not the other FO (i.e. the real one), it's very common in avaition to allow the most junior (in the seats) to fly during a problem, and the experienced pilot tries to work out what the aircraft systems are up to. Of course the implicit assumption in this is that a) that pilot actually knows how to fly, and won't make things worse (that was quite true until the advent of the most recent style of cadets), and b) the aircraft doesn't present immediate handling problems that need as much experience as possible applied to them. The upshot from that, is that you really should pick your mark, and have a plan for handling the 'who will fly/who will do the systems' question, based on the experience of the person you are sitting with. Contrary to what you may expect, if all other things are equal, the captain will generally let the FO fly, whilst he works with the systems, as it gives him a much better idea of just how to manage the problems. Flying with an SO on the other hand, requires more thought...something you do every time you meet a new crew member.

Do you think there might be of been problems with the control-stick or system behind it?
It looks to me like there were problems on both sides of the joystick.

If your question is, did the aircraft cause the accident, then I'd say the answer is no. It most certainly should not have had multiple airspeed indication failures...that's indicative of very poor design. But, loss of all airspeed data should not be hard to handle, much less lead to a crash. This has happened before though, a lot of times over the years (a 727 and a 757 come to mind...I think in those cases insects had blocked the probes). But, if the question is, did the aircraft make it hard, through confusing, and MOSTLY UNDOCUMENTED behaviour, then I'd have to say yes. They would have had multiple warnings, and probably not all immediately looked to be related. The ECAM would have been applying its priority logic to them, so it's quite possible that warnings appeared, stayed long enough to be half read, and then simply disappeared to be replaced by something else. Also possible that they flashed up long enough to distract, but without ever allowing them to be read.

But, ECAM would not have made the aircraft crash. It crashed because it was not being flown correctly. Problem is though, that in the modern world, quickly trained, minimally qualified (i.e. cheap) pilots, are part of a system that believes that the aircraft are 'uncrashable' as long as the pilots don't touch them. The entire Airbus philosophy is based upon the pilots fixing up the ECAM (i.e. systems), and letting ze computer fly the aircraft, as it knows best. The problem is that the computer gives up the ghost at the very slightest provocation, and when it does, it invariably drops an aircraft that is more difficult to fly into the pilots' laps. And, because of the changes to control laws, it may not even fly like it normally does.

As for why it stayed stalled...well firstly, I'm not convinced it would have been all that easy to unstall it, given the position of the tailplane. I don't know the A330, but in other Airbus types, the switch that moves it is inhibited (i.e. doesn't work) in flight, UNLESS the aircraft is in direct law, and there's no mention of that on the released info. There is a way of forcing the aircraft to direct law, but no mention of that here either. Nevertheless, even if it was having trouble overcoming the tail, you most certainly would have expected to seen full nose down joystick application. So why not? Well, what it appears happens is that as the airspeed indication slows, once it gets slow enough, the system rejects it. So what does that mean? Well, it means that it removes the information from the displays entirely...so even if correct, the pilots aren't shown this data. Secondly, though, the angle of attack system provides the stall warnings. And it does...until the AoA figure becomes so high, that the system does not believe that either, and rejects it too...and at that point the stall warnings cease, because the aircraft has rejected the data that says it is in one. So, pilot applies some forward stick, AoA reduces, and lo and behold, the stall warning starts again. He literally has to push through quite a period of warnings, before he would be able to get it to validly stop. And of course, he never did.

The pitch attitude, and performance tells you that you MUST be stalled, but the aircraft systems (and ze cannot be wrong) can't make up their mind.

Still want cadets in airliners?
 
Actually, after reading that, I don't want Airbuses as airliners!
 
The pitch attitude, and performance tells you that you MUST be stalled, but the aircraft systems (and ze cannot be wrong) can't make up their mind.

Scary stuff when barreling along at altitude. (or downwards in this case)
 
Thanks for the analysis jb747.

So, apparently, the aircraft was actually being flown, at least for much of the event, by the most junior of the pilots. More of less the equivalent of the 200 hour cadets that Jetstar thinks is such a good idea. He may have had more flight hours, but in all honesty, you do not learn how to fly in the coughpit of an airliner, nor in a simulator. You do learn how to manage systems, but, not how to fly.

Yep - keeps sounding more and more like Airframe. :-|

Airframe (novel) - Wikipedia, the free encyclopedia
 
From what I've heard the aircraft after stalling (+40 AoA IIRC) started to fall tail first and consequently crashed into the water in that position?

I stand to be corrected, however I would think that'd explain why recovery was impossible given the lack of air over the elevators, therefore making them useless and unable to provide aft lift to push the nose down. I'd imagine the same would go for roll. Also if the aircraft was indeed falling backwards wouldn't that significantly degrade engine thrust? My logic being sufficient oxygen would be unable to enter the engine to provide the necessary combustion required for maximum thrust in conjunction with the long spool up time of turbofans?
 
From what I've heard the aircraft after stalling (+40 AoA IIRC) started to fall tail first and consequently crashed into the water in that position?

I stand to be corrected, however I would think that'd explain why recovery was impossible given the lack of air over the elevators, therefore making them useless and unable to provide aft lift to push the nose down. I'd imagine the same would go for roll. Also if the aircraft was indeed falling backwards wouldn't that significantly degrade engine thrust? My logic being sufficient oxygen would be unable to enter the engine to provide the necessary combustion required for maximum thrust in conjunction with the long spool up time of turbofans?

No, it was NOT falling backward. Look at the numbers. About 16 degrees nose high, with an impact speed of 107 kts. Roughly, it was going forward at about the same rate it was going down.

The tail plane was not stalled, as it was about 15 degrees nose up. But the tailplane was also set to about 15 degrees nose up (which means that it was tilted nose down when you look at it). Actually, if the tail had stalled, the force providing the 'nose up' input would most likely have disappeared, and the pitch would have reduced.

The sort of thing you are talking about is a 'tail slide'. Quite a nice party trick in an aircraft that can do it, but basically self recovering. The aircraft needs to be in the absolute vertical to start with, and will pitch itself to vertically down within a couple of seconds. Of course, it was fun in something like an MB326, but the forces involved in the correction would be well beyond what an airliner could cope with.

You can hold a swept/delta winged airliner in a stalled condition. Not something practiced in the civilian world, but the older fighters spent lots of time 'ballistic'. If you select the right g, the aircraft won't be stalled even at very low speeds. The behaviour isn't necessarily what you'd expect from flying a Cessna.

The engines are neither here nor there in this event. TOGA power at altitude is virtually exactly what you were cruising with anyway (put simply, they run out of power at altitude). Selecting them to idle should be helpful if you're not getting the pitch response you want, as they normally provide a nose up couple, so idle should help get the nose down.
 
I don't recall any media story about the plane falling backwards into the ocean, but I am puzzled by this last paragraph in The Sunday Times article that was reproduced in The Australian :-

"In the final chaotic moments, both Bonin and his co-pilot attempted to simultaneously operate their side-sticks, before the 32-year-old seemed to give up. “Go ahead, you have the controls,” were his last words, possibly directed to the captain, who had woken up and rushed into the coughpit, but was too late to do anything."

According to the interim report the Captain re-entered the coughpit at 2:11:40 with the Airbus at around 35,000 feet, and the plane hit the water at 2:14:28 which was nearly 3 minutes later. He had only been out of the coughpit for 10 minutes so I can't imagine he had been asleep, which makes his apparent lack of intervention more puzzling. How did he also not realize that the plane was dropping like a stone with its nose high? Isn't there an artificial horizon that would indicate this?
 
Status
Not open for further replies.

Become an AFF member!

Join Australian Frequent Flyer (AFF) for free and unlock insider tips, exclusive deals, and global meetups with 65,000+ frequent flyers.

AFF members can also access our Frequent Flyer Training courses, and upgrade to Fast-track your way to expert traveller status and unlock even more exclusive discounts!

AFF forum abbreviations

Wondering about Y, J or any of the other abbreviations used on our forum?

Check out our guide to common AFF acronyms & abbreviations.
Back
Top