When you learn to fly by instruments you need to ignore what your body feels. Vertigo will kill you if you let it. JFK JR did just that when he went in his graveyard spiral. I have been there and know what it feels like but when it happened I knew if I didn't believe those instruments and went by my instincts I was dead. I was laying over half sideways fighting the instinct to go with my body which felt upright and that self survival instinct vs. believe the instruments. We were IFR in a Lear Jet at high altitude and didn't have a lot of time to sort things out. Pilot experience actually flying, not monitoring an autopilot, sometimes pays off. It would be interesting how much stick time these two had. Monitoring the autopilot doesn't count.

My buddy was flying the empty Lear Jet and decided he wanted to hand fly it while I was balancing fuel with my head sideways when I saw everything going bad and took over. It ended up ok but could have been a disaster for us. I couldn't sit up straight until we descended below the clouds.

Seems like if the PNF had been more agressive he could have prevented what the PF caused by holding full back on the SS.

Just a wild stab at "fuzzy" logic from non-professional:

They were going down and they could not understand what is going on for an extended period of time. That can mean that by that time they also ceased to trust some of their instruments. So what did they believe from the data they were seeing or assuming? That can be deduced from the only constant in their behavior: "stick held fully back" most of the time. In any dangerous situation with unreliable information present, you are relying only on those information that you assume to be most reliable, and you are doing only what is most important for the moment. You will stop falling that way only if you have forward speed, so that must be what they thought to be most reliable of the important information. It is very difficult to assume that all THREE pilots could miss the very possibility of stall, regardless of their training and practical experience. One, yes, but all three, no. However, if you are sure that you have forward speed, there is no stall, so that neatly explains both facts. If it is really true that they tried other things, but stubbornly insisted without changing only on that one action, than that is what all three of them believed.

Once stabilised, there is no 'acceleration', and one does not feel velocity, but one can hear it. There were no CUES, the STALL was abnormal. They felt more or less as they would have felt on the ramp. There were fluctuations in VS, and these would have added to the confusion, imho. Accelerating from the top of climb would have likely been exhilarating, but reaching 100 knots down in vertical velocity would not take long. The ship showed manageable, if not pleasant characteristics (in the descent, STALLED), again a confusion in territory one has been led to believe is mysterious, and unavailable to the BUS.

When stable at 1g, the Vertical speed would not be believable, it would be hard to accept. Close to the surface, the pilots comments were limited to specific altitudes, not their vertical speed. Again, they had no instrumentation or feel that suggested their predicament. No "believable instrumentation", that is. So get a grip, and lose the feigned shock; in that cockpit, no one would know what was happening. Acting like these three were too thick to sense the obvious is ridiculous. Nothing was 'obvious' and I feel no flyer would have had a clue.

What mystifies me is the power they carried throughout. If too fast, and in some kind of descent, closing the levers would be second nature. Is there a remote possibility They were in Autothrottle? Could they have been in Autopilot? At any time after the first loss of it?

Have BEA rejected with proof that the climb was uncommanded, or partially so?

one of.... if not ...the best statement so far

I'm not sure if this has been asked or discussed previously, but do the pilots have a GPS or INS based groundspeed readout on their displays? Also i was under the impression that heavy jets are equipped with AoA sensors?

Would not a very low groundspeed and high AoA readout cause the assumption of a stall? especially seeing as these readouts would not be related to the Pitot system if it was faulty.

Sorry if this is unrelated or has been discussed I have not followed this from the beginning and have not flown anything more complex than piston twins.

One G ?
 

Lyman, this one G thing confuses me. One G is what you get when your NOT falling, ie what the bus uses to fly level I've read, and what you'd feel on your bum. I'd have thought that falling at their rate you would "weigh" a fair bit less all the way down. Not suggesting that would matter in all the confusion, but physics interests me.

true...but only during acceleration
after certain speed is achieved = the drag and weight are same, no more accelaration and you have a pure 1 g

Wait/ Weight
 

I'm sure your right but I'm talking about the people in the a/c. Does that work for them? I'm thinking about the astronauts free falling in a jet nosing over, or astronauts in orbit, speed constant yet no 1g for them.

The astronauts are weightless only while the aircraft is 'bunting' over, accelerating downwards, and pushing negative 'g'.

In orbit, the vehicle is constantly accelerating downwards (a process offset by its speed).

No need to go in a aircraft for experiment sensations of vertical displacement
Try an elevator (if never experienced !) and you will understand .. and it's cheapest than an aircraft experience :ok:

jcjeant: spot on. Air Tourer, I concur, the elevator test is good. I tested that on a ground to 35th floor in a business tower, because I was amazed by how "smooth" the ride was, however quick. I used my smartphone accel sensor to check how many g I "took" in this elevator.
~1.15 max when accelerating from ground level or descelerating arriving at ground
~0.85 min when descelerating at top or accelerating from top to ground
1 g stable during the stabilised phase of the "ride" (=no "falling" nor "going up feeling", except for the ears) ;)

About the sensation by the ears: nope, the plane was pressurized, i.e. constant cabin pressure above 8000ft ;)

Please forgive what may well be a crass comment from a non-pilot who lacks the benefit of those hundreds of hours of professional training.

It would appear the crew were in "unusual attitude " situation.

Logic, to me, dictates that, if the aircraft was in the cruise prior to the upset, the trim and power were set for stable flight.

It would seem elementary, that no matter how much increased power was applied, continued back-pressure on the controls must result in either a loop (repeated until the stick is returned to the "balanced flight" position) OR the angle of climb would increase until the power available balanced gravity(+ lift) at which point the thing drops out of the sky ,tail down, presumably with reduced thrust available due to the abnormal airflow around and through the engines. (effectively a high-speed stall?)


It's easy to sit pontificating, but it seems they forgot the fundamental physics of flight and sat in a blind funk of panic.

No, I'm not a "simmer" :8

This wouldn't happen while stall protection is still available (which as far as I'm aware it was). I think it's far too simple to assume that the PF had, for some reason, temporarily forgotten the physics of flight. Far more likely to be a combination of factors including (though not limited to); severe spatial disorientation, confusing messages from the a/c and possibly the belief that pulling back on the stick was the best way to recover that particular a/c from the situation he believed it to be in.

I'm amazed at the amount of "needless" speculation for a relatively simple event. The PF simply failed to respond correctly to what should have been a non event. Had he followed his training and flown pitch and power then this is just one of dozens of similar incidents that plagued the 330/340 series.

His error was compounded by the senior pilots (on flight deck) failure to exercise command authority even though he knew the PF was over-matched by the situation.

I do not believe the aircraft was recoverable by the time the Captain returned to the flight deck however I do believe his failure to assume command (as PF) was an incorrect decision.

There is no indication that the initial climb was anything other then a response to stick inputs by the PF. Further the trim was a response to his continued application of control input. There was no autopilot or system induced "runaway trim". The aircraft was perfectly flyable from the moment AP disconnected until it hit the water.

At the end of the day the lesson is simple...the more you en devour to remove the aviator from aviation the more likely you are to have this type of end result.

g=1 when falling with no acceleration.
 

You feel weightless only when you're accelerating down. When you're falling at a steady speed you feel 1g.

The elevator comment is absolutely right: when cruising between floors, you can't tell you're moving. The light and heavy gut feeling only comes when the elevator speeds up or slows down.

Astronauts are free-falling continuously around the earth: the earth is just curving away so they don't hit it. They are accelerating towards the centre of the earth under gravity at 1g, but so is the space station, so they feel neither weight not a reaction force from the walls.

If you're falling steadily, with weight balanced by drag, and so no acceleration - either when `free-fall' parachuting at terminal velocity, or aboard AF447 - your gut feels like it would sitting on the couch.

Added:

Although you would feel like you were sitting on the couch, it would be a couch tilted back against the wall.

A plumbline would show your attitude, although you'd need a hanging spring balance to confirm there was no acceleration to confuse the result. There may also have been enough shaking to make that ineffective.

If the captain had clung to a handle rather than strapping into a jumpseat, that might have been enough to convince him of the nature of the situation, as his feet would have dragged aft compared with his head to balance.

ok here goes, virgin poster so go easy on the flames, eh ...

As i understand it, this aircraft was descending "straight? and level" at 10,000 ft/min therefore its AoA was pretty much 90º. Would it have made a blind bit of difference if the stick was fully forward in this situation as the air going over the elevator wasnt coming from an influencable direction. -- i just made that word up !!!

When i read further in the thread that having low slung engines prouduces a pitching up affect with power, then "stick forward and full power" would seem, to my little mind anyway, have bugger all chance of saving the souls in that situation anyway.

Power at idle and "this is your captain speaking, could everyone onboard run like hell up to the front of the aircraft" ... might have had more chance of getting the thing back to controlled flight.

mind... 1st post so go easy.

Ahem... no. Incorrect assumption. "straight and level" doesn't imply "with zéro forward speed". Only if you have "straight and level + zéro forward speed" will your AoA be ~90°
Actual AoA from AF447 can be read easily on the published FDR traces, it was never close to 90°.

However:
- given the situation, indeed TOGA was bad, Idle would have been better
- given the situation, indeed all pax in front would have help; no idea if it would have been significant

Re: 1g felt?

The BEA interim report #3 shows the last data before impact of 1.012 g Nz, -.17 Nx (lateral), -.10 Ny (longitudinal). 16.2 deg nose up, 5.3 deg L roll, -10912 fpm. The traces show Nz typically between 0.9 and 1.1, with occasional increases to around 1.2, during about the last 2 1/2 minutes of flight.
Edit: Ground speed 107 kts, so it wasn't close to a 90 deg descent.

So, yes, the descent was in a pretty much a "1g environment".

ok, ma bad. still trying to get my head round how it could have had forward speed with a nose up attitude and that rate of sustained descent. however - thats my head and my problem.

Heck - I just keep on seeing that bird..................................anyone got a 12-bore?

AOA above stall doesn't matter much. No it wasn't 90 degrees but it was held above stall by the PF for over 3 minutes. That is what caused the crash and the PNF not intervening. 1 G is a constant like going down an elevator, descent rate doesn't matter once established. The only way they had to save themselves once in the descent was to lower the nose if possible so they could be unstalled to generate the G's to recover. No one did that.

Why? You learn that in the first 5 hrs of flt training.

Pitch and power isn't the only thing one is trained in (granted, it's a core fundemantal)

Was what his training?
What was his currency?
What else was he trained to do?
Did he receive any negative training?
Why did his instrument scan break down to soon?
What element of his training, or lack of it, or lack of recency in training, factored into that?
(I am making an assumption that the PF's instrument scan broke down early in the event, as evidenced by his climbing 3K or so without ample correction well before alt change went that far. If my assumption is off, so be it.)

Stall protections (in fact all hard protections) were gone as soon as the speed went unreliable. Alternate 2 (NO PROT) means just that.

High altitude stalls
 

My last 777 sim session in July, I specifically requested we experience a high altitude stall. In my 33 years of commercial flying and 12 years of US Navy aviation, I could not remember ever seeing a high altitude stall demonstration.

We ran the sim up to 37,000 and disabled the automation (auto throttles, auto pilot). Pulled the power off and waited. When the aircraft stalled it entered into a rapid descent, nose up attitude with some wing rocking. We first applied max power without lowering the nose; no recovery. Aircraft did not exit a stalled condition until we agressively shoved the yoke forward, reduced AOA, and regained flying speed.

It sure made a believer out of me. If you stall at higher altitudes (above 30,000), it will take an agressive push over to recover.

We all practice stalls at lower altitudes in training where all it takes is the addition of power and a slight lowering of the nose to recover.

Next time you are in the sim, make your instructor demonstrate this.

Clear explanation why accurate air speed indication was essential. According to this, margin of error was only ~ 25 knots.

"As the speed of sound depends on the temperature of the air, which decreases with altitude up to the tropopause at around 17,000 metres, the maximum speed an aircraft can fly at safely (before running into speed-of-sound effects) likewise decreases with altitude. Thus, on a chart of altitude (vertical) versus speed (horizontal), there is a point where the positively sloping plot of the plane’s stall speed crosses the negatively sloping line of its maximum safe speed below the speed of sound (Mach 0.86, in the case of an Airbus 330). The apex where the two lines intersect—where the minimum and maximum safe speeds are the same—is known euphemistically as "coffin corner". At 10,600 metres, a fully loaded Airbus 330 cruises (for reasons of fuel economy) just below this critical point in its flight envelope—with probably no more than 25 knots (46 kph) between stalling (through flying too slow) and breaking up in a shockwave-induced dive (through flying too fast).
....
With the plane buffeted violently by the turbulent updrafts, the first thing the pilot would have done would have been to reduce speed to ride out the storm. But with no clear idea of what the actual airspeed was—and with no automatic systems to prevent manual inputs from making the aircraft fly too slowly or too fast when operating so close to its coffin corner—the pilot could have unwittingly pitched the plane into an uncontrollable stall."

Aviation accidents: The Difference Engine: Wild blue coffin corner | The Economist

Jjeppson, can I ask at what altitude you were when you managed to exit the stall in the sim?

"We all practice stalls at lower altitudes in training where all it takes is the addition of power and a slight lowering of the nose to recover."

More likely this is, stall warning practice or 'approach to stall'.
1. Stall warning = Approaching stall - Actions are required to avoid stall.
2. Stalled = Stalled - Actions are required to recover from stall.
3. Deep Stall = Stalled and no Pitch Authority - Recovery unlikely, at least by using elevator.

Some have suggested, or implied, that 'Deep Stall' may be defined as 'very much stalled', but this offers very little additional information than 'Stalled'. Which is why we refer to a Stall which has also removed normal pitch authority as a Deep Stall.

I'm curious if there is any thought that the PF felt that his SS input would have the effect of having the computer fly "pitch & power"? That he was still somehow "protected"? In the absence of such a train of thought I can't comprehend how he would fly "up"...knowing that every meter of altitude gained reduced his margin of error with regard to speed. It would seem that if anything the natural response would be to move lower to widen the speed envelope.

Psychologically, I agree. However, law change should have alerted him to lost protections.

The PNF, once again features, based on CVR audio. HE appears to have sussed the situation out, but did not assert his observations.

This is going 'round in circles. However, the PNF took control for a few seconds and the PF re-assumed control in an unethical manner.

This, to me, smacks of an ego clash - remember, the PF's wife was onboard.

I agree (from what we know) that the captain did not assign authority accurately.

I suspect that when the aircraft initially stalled, the 1g acceleration earthwards would have created a "falling sensation". However once the aircraft reached terminal velocity downwards there is no further acceleration just a downwards velocity vector. Then there is no acceleration hence no "falling" sensation anymore. How long it took to reach terminal velocity, no idea. Someone cleverer than me can probably calculate it with the right data. I'm guessing the Captain rushed back up front when he felt the vertical acceleration before terminal velocity....

From a blog.

They did not know the a/c had Stalled. They did not have Atttitude displays on the panel. Had there been any attitude displayed, anywhere, there would gave been a reaction on the part of this crew . This is so blindingly obvious, one cannot fathom the continuation of all the drivel re: PPL, ab initio, etc.

caveat. If they had an attitude display, it provoked no comment, no reaction, and no solution, nor discussion of same. In other words, it fit their conversation; it was roughly level, or slightly descending. To assume these guys had any provocation of the STALL shows only the ignorance of the critique. As Brian's quote above, if these guys did this, and the posters here are correct, we'd all best be not flying, anywhere.

To continue with the incredulity shows how eager are some to have a quick solution, a reason, a soothing, of the Fear. In the face of common sense, can you not see this? Their behaviour is readily explained, and if one took the time, carrying an open mind, the evidence fits.

Play it any way you like, there are no rational explanations except these: these gents were trained, they were alert, and they were together.

They had no clue, because there WERE NO CLUES. Resting one's lazy arse on data that got pulled out of the drink, data that they had no access to, To condemn, is not just lazy, it is a mockery of itself. It is right in front of you, it is not mysterious. More, much more than they had.

Stop leaving unsaid that which you pretend not to know: Look at the reasons, the logical ones, not the snarky gossippy drivel about how the training must be at fault.

You are looking at data that did not exist in a real way for this flight. It is a time capsule, and unfortunately it is being believed, and used, as if its purpose alone is to lend credence to a slander.

Final "conversation", and we all read it, comment it, teach our little lesson...

1 g earthwards is what you're feeling right now.
Or he could look at page 106 of interim report #3.

Oozlum time again. - which does NOT apply to AF447, and as far as we know only applies to T-tailed a/c (OK - or perhaps one with excessive tailplane icing) - just to make sure we do not set off yet again on one of our circles. Are we clear?

PS Forget these red herrings like the NASA Schweizer in #431 - two minutes reading up on the project would show it to be interesting but irrelevant.

[quibbling]Not exactly. g protection remained; which is a good thing IMO, should a recovery attempt have been made after the stall. [/quibbling] ;)


jjeppson:
Thanks, that shows (at least) the A330 & 777 sims perform the same. Until proved incorrect, I'll assume the planes themselves do that too.

Lyman:
Until proven incorrect, I'll believe they had three or (at last) one attitude display. They are mentionned in the CVR. "Selon les trois tu montes" could be about the 3 attitude displays or the 3 V/S or a combination of attitude+V/S+altimeter. Later, "l'horizon secours" is mentionned by the captain...

While we are going over old ground - can anyone give me a link to an explanation of why PF's 'instruments' were placed in ATT by PNF (2:10:39) and what it means? Why would it have been done - is it a QRH action, and if so for what?

My contribution to the Oozlum flying display:sad:

My apologies, I stand corrected. I think the PF could have made the same mistake as me!

I'm neither a newbie, nor a greybeard, but have only experienced Boeings, and admit to healthy distrust, due to my Luddite tendencies, of Airbus.

What astounds me is that this pilot held the stick back for three minutes, almost as if he'd lost hope, and that was all he could think to do. It smacks of someone continually pushing the lift button, harder and harder, even though the lift has been called.

I find it very difficult to believe that any pilot of my 737 would sit there, for three minutes, with the yoke in his stomach, the stick shaker rattling, and the altimeter unwinding. Boeings have their faults, and all automation has its traps, but this accident does nothing to quell my Luddite fears of Airbus.

It's all very well saying that you can fly an Airbus, it's just like any other aircraft, but it patently isn't. It's been designed to be flown through the autopilot, and seems to do that very well, and provide a lovely working environment. However, the downside is the pilot-machine interface. The sidestick and stationary thrust levers offer no tactile feedback to the pilot, as to what his aircraft is attempting to do, nor even any clue to his colleague when flying manually.

The Airbus sounds lovely on a normal at the office, but seems to turn into a nightmare when things go wrong, and the very things that make it work so well on a normal day, seem to make the pilots even less well equipped to deal with problems when they arise.

The pilots were not linked in. They were sat there scratching their heads, basically. All they had to do was fly the bloody thing, but it didn't seem to occur to them.

Slickster, that is all so true. Modern aircraft seem to be full of nasty technological traps.

The 777 has a Thrust Asymmetry Compensator (TAC), which puts the rudder in for you with an engine failure. While I was there, every other Base Check was performed with the TAC working. Easy peasy. But you only had one go per year of feeding in the rudder manually.

I don't recall who the operator was, but one of the first real engine failures on the 777 sent a big chunk of metal through the cowling, and straight into the TAC, taking it out!

The PH coped ok, but then he was probably (in those days) an old stick n' rudder guy.

Here's a case where the one day you need stall protection, it ain't working!