A flight deck is no place to rely on the amygdala, least ways not in cruise.

It is said that it is a good bet that those of us alive today had ancestors who were good at spotting movement, not detail. Detail is for which part of the animal we wish to consume. Noticing movement keeps us out of the sabre tooth's maw.

Left Brain digital; good for ennui producing comfort, the time after the hunt, and for making spears. The ever ready amygdala wears us out, and makes us quick prey for the threat we don't see, being exhausted from hyper vigilance.

The computer has no skin in the game, save for the Hubris of those who built it, sold it, and equip there a/c with it.

Complex is for rules. Simple is for survival. We have known that for thousands of generations, yet we convince ourselves it is 'different' now, we have 'evolved'.

Of the hundreds of times I have stalled a flying a/c, not one time was it inadvertent. Once on approach, a traffic conflict got too much of my attention, and I got slow. I started a turn, and heard the horn. I am happy to say I did not have to think, consult a manual, or radio for assistance.

Nothing like the hangman's noose to focus one's attention.

This path started by Mimpe is very germane. Google University of California, The Mind Institute. Wander into the Hall of "Brain Mapping".


Just to put a little anxiety into our fbw partisans. I will wager that prior to ai flying a plane, the brain will be connected through electrodes to a/c controls, and said brain will fly that a/c in ways that rules based ai could never accomplish. Case a Guinness.

Remember in the mid-eighties that a lot of performance cars were going to video display full digital dashboards with no analog gages????

Then the guys actually running cars on the track figured out that good old analog gages were actually better. They found the driver didn't have to actually look at the gages to know all was ok and what the car was doing. The information was received and processed by the subconscious within the normal scan of the instruments because the regular positions of the needles were within the driver’s memory. Not sure that tapes can give the same situational awareness as good old fashioned gages.

I'm sure modern glass cockpits are much better than steam gages, but maybe there are some things like altitude, airspeed and vertical speed that might be better presented by a gage.

A little stocktaking is now required ... well said bear.:ok:

Ears, Brains and Seats
 

Speaking of brains and our human response to various stimuli -- computer aided or not -- what would the "seat of the pants" part of the pilots' brains be experiencing?

In this type of stall, would they feel increased positive G's which would then seem to decrease (or even go negative) as they rapidly descend in the continued stall? I gather the BEA will provide a graph with G forces plotted with a full report of FDR data, but no doubt it could influence the pilot's actions or lack thereof.

KMD: That is one of the reasons floating point operations are avoided in control systems. Other representations/encodings are invented.

BreezyDC
The following applies to the developed steady state stall:
If they were not accelerating or decelerating and the wings were level, and the nose was 16 degrees in the air, then they would feel like they were in a slight climb attitude, but without that push in the back that says you are accelerating.

When a wing drops in the stall, the wing is still generating some lift which can turn the aircraft, but since the aircraft fuselage is also now being supported somewhat by the uprushing air, there will be much more lateral acceleration (downward) than you customarily experience in a wing down attitude.

When in doubt remember, all forces have to be in balance or the aircraft would be accelerating.

As all you would know, their seat of the pants would be

a) screaming at them when they shouldn't or
b) not screaming at them when they should
c) or ( most probably ) both at once


and most of the resulting spatial inferences in this situation would be incorrect and dangerous

engine-eer
I would like to point out, that "better" has different modes of comparison.
Price? Cost of ownership? Reliability? Ease of maintenance? Intelligibility?
More easily integrated into the instrument panel? More rapidly assimilated?
ETC.

It is all a tradeoff. The issue is how to cover the maximum number of bases, and still give the crew what they need in the worst case they may face.

A mix of old style (steam gage) and new instruments (digital) can best do this.

Another hi alt stall Report: West Caribbean MD82 at Machiquez on Aug 16th 2005, did not recover from high altitude stall

G's
 

Here is a graph showing my estimate of the G's that produce the trajectory and speeds described in BEA's Update.

to Harryman

'Best read the few thousand previous posts on all AF447 threads to catch up a bit first, then maybe post.'

Thankyou for your slightly patronising tone! I have already read all posts. If you think I have said something dumb, please could you explain - I would prefer to be educated than mocked ! thanks, paul

HN39, am I reading your graph correctly ? Shortly after 2:10:30 the a/c started to descend (g less than 1.0) at which time the pitch was greater than 10 degrees (and increasing) and the FPA was about 8 degrees and just about to max ?

Presume in a stall vertical acceleartion (g) is less than 1 and FPA will decrease, so are there two stalls, each starting where FPA starts to decrease (and g about 1 and falling ?).

No, you are not reading the graph correctly. Perhaps you should read it together with another graph that shows vertical speed and altitude.

Per BEA the airplane stalls around 2:11:00, after AoA exceeds alphamax and then approx. 9 degrees and continues increasing as shown in the first graph. At 2:10:30 the vertical acceleration reduces to less than 1g. Vertical acceleration above/below 1g is the rate of change of vertical speed, which is 7000 fpm at that point and starts reducing to 700 fpm.

Agree.

I long ago threw away one of those new-fangled Seiko World Time digital watches, one looks at The Picture on a normal watch, don't have to 'read' and assimilate the information.

Mind you, I then went for one of those Glycine jobs where the hour hand goes around only once in 24 hours, so that 12 Noon is at the bottom - that threw me for awhile !

There is one way to settle this once and for all.

Get Airbus test crews to take up an A330, with the same w & b and put the thing into a deep stall. Ballast could be gleaned from the sales team. The data from the result could be feed into all the A330 Sims and it could benefit all.

They wouldn't need parachutes as the AB is uncrashable I am told as the built in protections and proper pilot's would guard against this..

HN39, OK, thanks. Will have a bigger think.

doubleu-anker
Doubleu-anker, I know your are being a bit facetious, but the core of the idea has merit.
Suppose we took a life-limit A330 or a ferryable but badly damaged aircraft, mounted high time engines, installed an ejection seat for the test pilot (singular), installed a ballast control system so you could move the c.g. around quickly, installed high speed telemetry, and went flying.
You could fill all the high risk niches of the A330 performance curves with relatively low human and financial risk. Might even collect enough data in one flight to do the job.
And when you get done you will have one sweaty test pilot who is an expert in flying the A330 at any angle of attack.:}

Yes, that's what these arrogant frogs were trying to make people and airline managers believe.

This arrogance was payed with blood pretty soon and some airlines made up their mind to improve training and not to reduce it, like the smart AB managers suggested.

HN39, I looked at the second graph, and see that toward the right side, there are vertical velocities up to 13,000 fpm. I gather that you have integrated the whole postulated flight path so that the end time works out the same as the real deal.

I also interpret all the acceleration figures to be relative to the surface of the earth. Have you worked out the perceived accelerations relative to the aircraft axis? That would give us the seat of the pants feel, or have you already made this transformation?

Another autotrim kill
 

MOTOROLA
Thanks for that report. I hadn't seen it before and it is relevant. The crew in the linked report tried to get more performance from the aircraft than they should, but autotrim is what led them down the primrose path.

I wonder if we still had the trim buttons on the stick/wheel whether we would be having this type of accident?

Yes, trim runaways seem to be prevented with properly designed autotrim, but the surreptitious changing of trim without annunciation is clearly hazardous.
At the minimum, the Airbus needs some sort of loud clicker when it changes trim so that the crew is aware of what it is doing.

A whole string of loud clicks would be a warning to see what is going on.
But with the AF447 commotion in the cockpit, would it have been audible?
Over to you, Airbus, for some redesign work.

I thought "traditionally" the pitch trim wheel went "ping, ping, ping" like a bicycle bell?
Easily distinguished from just about everything else.

Machinbird;

First three sentences: Yes. On the other hand, I was mainly interested in the motion of the airplane, and did not work out the "seat of the pants feel". The lift coefficient and hence alpha is based on the acceleration normal to the flight path.My data do not go beyond alphamax, so anything beyond that should be taken with the appropriate grains of salt.

That's a depressing read. At least they got around to radio ATC for clearance to fall out of the sky :ooh:

(from Report: West Caribbean MD82 at Machiquez on)

Hi

to fill some emptiness (no job presently) I started to build a model into x-plane, a simuator game that claim to calculate air forces mostly correct.
I did a blended wing, during flight tests i got severe problems to recover high altitude stalls. I figured out a reliable procedure for x-plane to recover from such a stall.

First let my say all what is reported here about that behavior looks reasonable to me. The 7000ft/min compares perfect to the trim setting. (What is not fully correct the trimsetting is corret for the speed 60-80 IAS, but the vvi is looking good for the flight maneuver.)
The stall speed is independent from altitude regarding IAS as long no oblique shock or vortex occurs.

My problem was after the stall the ACFT reached a stable position with 20-30 degrees AoA what i not could not terminate. After many crashes I decided to fly the stall what mean not try to recover early by pressing the yoke forward. I pulled gently until the rear stop was reached, power idle, engine secured by igniter, while gamble to keep wings leveled, checked the horizontal trim settings (I found best was take off trimm working). When reached that carefully move the yoke forward to reach more than 30 degrees downwards but not more than 60 degrees. While dive let the speed increase, just a little below maneuver speed bring pitch to the horizon. Reset power.

This is not the procedure for reacting for a stall warning alarm, what still is just add power. It is for my little blended wing in x-plane. I never lost more than 10000 feet but thats better than an impact.

What the pilot might have felt
 

If they stalled it at a load factor of less than one g, that would result in a stall at a considerably lower speed. The Airbus Safety newsletter mentioned that buffetting would occurr as the aircraft approached stall.

But in this case, if the aircraft stalled at less than a G, the speed would be lower and there may not have been any or much buffetting. If that is the case, the PF may not have felt enough of a stall or break to make him think he had actually stalled it. Stall horns go off when you approach stall, not after you stall.

What very well could have happened is that as the stall warning started and the PF applied TOGA. The aircraft didn't buffet, shake or break into a conventional stall, it stopped flying and started falling at a lower speed.

Absent the break or buffett, when the stall horn stopped the PF likely thought that he flew out of the approach to stall and was flying again. All speculation, of course, but if that is what happened it would offer an explanation of what the PF thought and would explain a lot better his actions.

engine-eer,
 

Engine-eer,
Sounds like a very plausible scenario to me, i.e. non-recognition of the full stall at the start (after intial nose-down following recognized impending stall) for reasons stated. But what is so baffling is that the PF (and others) would persist in such a belief in the face of a rapid and continuing descent for 30K ft with the stick full aft and substantial nose-up attitude (even if credibility in IAS never returned). [Assumes all relevant data provided by BEA, skeptical icon ...]. Of course a continuing stall alarm would have helped greatly one assumes but really, at some point reality should have dawned, leading to forward stick +/- power, i.e. at least an attempt at recovery.

Sorry to bring the discussion back to this issue but for me it's the most troubling one in a way. Let's hope the CVR will reveal much more about state-of-mind of some or all involved.

<returns to piston gallery>

Trim wheel motion
 

Christian, for things Boeing, I believe you are correct.

A while ago on earlier threads, PJ2 briefed that the A330/340 trim wheel moved silently.

Don't know why they made that decision at Airbus, but I think they need to revisit it.

As far as I remember, same thing on Concorde (Sud-BAC, then Airbus).
I'll have to check, or ask friends to check.
If even to an ancient as myself that was "obvious", one does wonder why it was changed.

The A300-600 has a whooler if there is more than a few seconds trim operation but it is cancelled when the autopilot is engaged.

Hi,

I think pprune can thank the BEA to have generate as much traffic on their forums by making public only a few fragments of the CVR
It is as if BEA would cause an outbreak of rumors ... not the best effect on the public.
When we know that their goal was to silence the rumors appeared in several newspapers .. it seems quite successful. :8
In short .. BEA com in all its splendor
One may wonder how to make public the entire CVR (less passages not related to technic matters) would prevent the ongoing and the result of the investigation to proceed calmly and with the confidence of the public.

The public calms down and moves onto the next disaster du jour when a a paper or talking head explains a theory to them with authority (one voice).

Since PPRune never speaks with one voice the media only use PPRune to incite questions.

So let's just carry on with our multiple theories based on what scraps of fact are fed to us :}

jcjeant

Red section...what are you suggesting?
The transcript of the CVR needs to be complete. Otherwise how do we know AF447 maintained a sterile flight deck? The Buffalo incident is a good example of non-sterile environment to the point of distraction in my humble opinion. Lest those lessons never be learnt.
Transparency is the only way forwards. Otherwise who do you suggest should erase parts of the transcript, AF, BEA, French Judiciary...?

"Transparency is the only way forwards."

I agree. Nothing good will ever come from secrecy, whether it be from a group, organisation, authority or society.

A certain European country learnt it's lesson last century. Their equivalent of parliament now has a glass dome for a roof.

Trouble is, any advertising/PR guy will tell you that first impressions last a long time - and, in many peoples' minds, for ever......

Some quite respectable newspapers, for example, appear to have got the impression that the pilot caused the initial 'steep climb.' Even the BEA thing doesn't actually say that - indeed, in very 'spare' wording, it rather confirms the opposite, that the pilot countered the climb with 'nosedown control inputs' and restored the aeroplane to a situation of reasonable airspeed and a reasonable AoA:-

So we're really no further forward as to what caused the stall in the first place?

But BAE's repeated mentions of 'noseup inputs' later on appear to have put over an impression that the pilot caused the whole thing?

My own position is that the BAE would have been 'within its rights' not to publish the CVR this early in the investigation. Alternatively it could have published the whole of it. But just publishing a few isolated quotes is, in my view, plain wrong.

And its also clear that, from the rest of the text, with its descriptions of control movements, power settings etc., they've already analysed quite a lot of the FDR information too. Again, I'd have preferred 'all or nothing.'

If you look at the paragraph preceding your selective quote (timestamp 11s earlier), you will find:
I concur that we don't know enough about the nature of that input to conclude with certaincy it caused the climb, but there's nothing in that note saying it didn't. So far the chain of events NU input by PF -> climb -> ND input -> reduced climb is the least complicated (and thus probably a likely) way to connect the few dots we have.

Sounds to me that you're making exactly the same assumption in reverse.

The investigation is without a doubt proceeding calmly. I am sure 99.99% of the public don't care/are not interested/have confidence. This thread is mostly pointless speculation with most of it being complete nonsense written by people who no little about flying jet transport aircraft e.g.Sterile flight deck? What?

I read this thread for amusement and for the occasional gem like post #1208.

Zorin 75:

What's depressing is how easy the recovery could have been by trading altitude for speed. That requires putting to nose well down in such circumstances followed by a judicious application of power.

aterpster: what I found most striking in the MD82 event was that the nose trim kept going up ... similar to what AF447 has.

Nose drop to lower AoA. Speed up. Watch RoD decrease.

I've been watching the comments on round versus strip gauges and personally agree: the strips don't give me the same "feel" as round gauges. Thanks to mimpe for a bit of an explanation on that. :ok:

Quote:
That's a depressing read. At least they got around to radio ATC for clearance to fall out of the sky

The lack of a distress call (or actually any call) from AF447 has never been resolved. Wonder if this was a conscious decision ?