............:ugh:

AoA an other unusual conditions
 

Quote:

---

CJ@#1314: To me, the suggestion of adding an 'unfamiliar, new' instrument (requiring additional training, too) because, once upon a time, people were incapable of inferring the state of their aircraft from the available data, makes no sense.

---

Why not to present this (important) info when AoA nears safe limits? Could be triggered by "rate". Warning a fast closing in to a dangerous condition. Requiring immediate action.

Certainly not relevant (and not justifiable) in normal conditions.

Another indicator IMHO should be Real Time info on near simultaneous "failure" of critical (redundant) elements, e.g. Pitot's.

The redundancy implemented in the design was not made to simultaneous failure of it. And the crew should be IMMEDIATELY informed of this unusual condition even before a/c degradation.

Why not? Cost is not an issue. Just a simple SW algorithm.

AFAIK we are still relying in "sub heated" Pitot 's.

Christiaan
Probably not Christiaan. The trusting souls who don't bother to cross check their instrument readings definitely won't need one. The ones who have never been taught how to use one are unlikely to appreciate its value.

However, If the simulator sessions emphasize cross-checking of instrument readings and otherwise exercise the use of an AOA gauge, then there will be a significant proportion of airline pilots who eventually appreciate what AOA information can do for them and they will use it.

I don't think it's red at all, and I think there are a few similar ones in the vicinity.

Same argument could apply to AOA indicator, if they'd had one - would they have believed it pegged at 35deg ?

And the speeds were actually valid again as they were stalling - but did they have any indication sppeds were valid again, and would they have believed the (correct) speeds ?

Then consider, not only was the needle pegged, but it is on a narrow vertical display rather than old style gauge. Does that make it easier or harder to notice the transition (maybe in peripheral vision) ?

And last but not least, instrument displays are no longer anything like single signals going to single gauges, but rather a whole bunch of signals going into a chunk of software and coming out as a bunch of pixels. If some of those signals are so far outside normal flight that you are in a completely untested region, who knows what pixels that software will produce ? If there is a flaw, it could affect more than just VSI. Or put another way, that "needle" may not peg, it might wander off the VSI and corrupt some other part of the display.

What we have in the FDR traces are the input signals to the instruments. If it was steam gauges, I'd be happy to trust that the gauge was showing the same - invoking occam to say that we should look for simpler solutions than a failure of a gauge independent but simultaneous with other problems. The behaviour when the signal goes out of range (hits the stops) is also simple to test and be confident in. For glass / PDF I'm not so sure. Not at all.


Sadly, we'll never know exactly what they were looking at, but I would hope that at the very least the HF investigation is feeding the FDR signals into a sim (or the iron bird) to get as close as we can to what the instruments were showing. If we had video of that here, I have a suspicion it would be a bit more confusing than what we may be assuming from the FDR.

There is also the red/black over/underspeed indication theory that I think has been posted more than once. If it's true, and if I understand it right, it may be a serious user interface design failure.

Turning
 

A " Rate 1 Turn" used to be 3 Degrees per second. For AF447 to have turned 180 degrees in, say, three minutes, would suggest that the aircraft was banked throughout this descent. Would this have averaged more than 10 degrees of bank ?

What else is the "STALL, STALL, STALL!" aural warning? It occurs before the actual stall (as well as supposedly during it, different issue, that).

My suggestion and argument, entirely.

Not necessarily "sub-heated".... but still "sub-qualified".
AFAIK, the Thales AA and AB probes, just like the Goodrich (?) probes, all met the certification requirements. The real problem is that none of the certification requirements fully account for the conditions encountered by AF447, and during other UAS events.
I'm looking at this a bit from the engineering point of view.... "I can give you a better pitot, just tell me how to test it to prove it's 'better' and meets your requirements".
And so far, AFAIK, those environmental conditions, and certification tests (dating from way back) have still not been updated.
The UAS procedures are nothing more than a stopgap and a BandAid.
UAS should no more be possible.....

In normal flight, I'd agree, but for diagnosis and recovery once stalled, is the simple binary warning/shaker (or not) enough ?

If we assume that everyone recovers promptly (and without getting into a worse secondary stall) at approach-to-stall, then again, there would be no need - but recent history shows that that is a fatally bad assumption.

Use of AoA Indicator
 

Well, how did they use it in the Concorde? Why was it there?

In carrier operations, it is the primary speed indicator for approach to landing. At approach speeds, it is operating in its most accurate range and precise speed control is essential to a good carrier landing. Every aircraft that I flew, the AoA indicator was calibrated to be "on speed" at 1.3 Vstall in the approach configuration. No matter what your weight, you were at 1.3 Vstall when the AoA indicated "on speed".

AoA is not useful at cruise speeds because the variation of AoA with speed is much smaller than at approach speeds. However, it could have been useful to the crew of AF447 in understanding their situation had they been trained to understand it.

As far as incorporating it into the scan, it is just like any other instrument. In fact, it is an excellent cross check against the airspeed indicator.

Hi,

HazelNuts39
From the CVR extract BEA report N°3 .. the first calls about altitude are :

2 h 13 min 32 PF call out level one hundred
2 h 13 min 36 PF call out Nine thousand feet
And the last (astonishing) call out from the PF is at 4000 feet
2 h 14 min 05 PF Well we need to we are at four thousand feet

Bureaucracy x good Engineering
 

Hi,

IMHO this is RIDICULOUS. Just waiting for going directly from Normal Law to Murphy Law. :ECertification "principle" is not being properly used. The bureaucrats (of low qualification:}) are in control?

IMHO, just RIDICULOUS. Airlines and pilots simply victims of bureaucrats.

Who deliver the a/c to them? With approval from who?

The short duration of erratic data in most UAS incidents shows we are near the "required Watts"(*). So they are certainly SUB HEATED.

(*) For most of situations

Vanished info other than (not measured) AS
 

Quote:

---

IFF789@#1318: And last but not least, instrument displays are no longer anything like single signals going to single gauges, but rather a whole bunch of signals going into a chunk of software and coming out as a bunch of pixels. If some of those signals are so far outside normal flight that you are in a completely untested region, who knows what pixels that software will produce ? If there is a flaw, it could affect more than just VSI. Or put another way, that "needle" may not peg, it might wander off the VSI and corrupt some other part of the display.

What we have in the FDR traces are the input signals to the instruments. If it was steam gauges, I'd be happy to trust that the gauge was showing the same - invoking occam to say that we should look for simpler solutions than a failure of a gauge independent but simultaneous with other problems. The behaviour when the signal goes out of range (hits the stops) is also simple to test and be confident in. For glass / PDF I'm not so sure. Not at all.


Sadly, we'll never know exactly what they were looking at, but I would hope that at the very least the HF investigation is feeding the FDR signals into a sim (or the iron bird) to get as close as we can to what the instruments were showing. If we had video of that here, I have a suspicion it would be a bit more confusing than what we may be assuming from the FDR.

---

:ok:

FDR in this case may not allow a "complete understanding". Lacks important info inside the loop (a/c-PF)

AOA
 

Stall-aural warning is an on/off indication, AOA is also a trend indication. Just look at it, see the AOA increase and expect trouble if you dont stop it prior stall range. If STALL sounds, you can check on it how far you have penetrated into unknown territory and how drastic your steps have to be to get out again. And you can crossscheck on it how effective the implemented measures are. You can also identify a false stall warning, which had been present in other cases with UAS and act accordingly.
It is not of interest how helpfull it would be when AOA is above 35°, as it should help to avoid STALL in the first place.

In an air transport aircraft it would be best used as a backup instrument in case of UAS (just fly a fixed AOA value and you are safe, regardless if you are level, climbing or descending) or other situations, where you are operating at the edge of the flight envelope.

The training you would need for that one is nearly none, you just have to be able to read the numbers. The AOA values of interest can be marked on the gauge by colours, those would be cruise, stall, max range, max endurance and others deemed necessary. No magic there.

AOA is a performance indication, and if it is an old analoge one i myself and most fighterpilots where used to, it is higly independent from the rest of the systems and electronic gadgets.

You might have it for years and only register its presence without using it in the daily normal work, but in an abnormal situation it might safe your day.

If it gets implemented, and i´m sure it will, there will be no complaint about it and no problem to make use of it.

Hi RetiredF4

Like this one?
Airbus backs up speed and altitude displays

I was used to a round one. AOA

The BUSS is again a cocktail of information with different sources and therefore inherent failure and error possibilities.

I would add it completely separate of the other stuff, a stand alone on if you like.

I have no definite answer on "should we have an AoA meter in every liner's cockpit ?"
I don't think "just" adding an AoA meter is a good idea. AF447's crew apparently didn't react on instrumentations they should have known/been trained for (namely: stall warning)...
AoA meter + associated training: why not. I agree that this training is simple.
AoA meter alone: not useful, potentially dangerous.

However:
Huh? How is that?
Stall warning logic is : ring if AoA > limit.
Then a false stall warning implies a false AoA measure IMO.
And a false AoA measure gives a false AoA reading on the AoA meter.
Same source = you cannot cross check.

Excellent post RetiredF4.

Many attitude display systems incorporate a PLI (pitch limit indicator) that correlates attitude with the onset of the 'stall warning'. This type of display could be utilized in alternate/direct law in conjunction with AOA.

Exactly. Stand alone or incorporated into an existing display thru the OPC (like the 737 NG PFD & HUD), data for AOA & PLI displays are already there.

First, i dont think that the STALL warning is derived from one or all three AOA vanes and not being run through some gizmo until finally used together with some other inputs like speed, loadfactor. configuration, altitude, you name it.

Therefore i have a more simple device in mind like posted above.

It would be most apreciated, if a backup system would not use the same source as the primary system. Schould not be rocket science.

If STALL Training is limited to "Approach to STALL" recovery, an AoA is a must?

but if you show the history of the datapixels on the display, mayby the least 60 sec, like in the datarecorder.... than there is a chance to trust the red line up to 2:11:45 and misstrust it later....
http://i45.servimg.com/u/f45/11/75/17/84/af447_25.png
http://www.servimg.com/image_preview...122&u=11751784

I did not keep up with this accident until after the recorders were found and the BEA made it's report this last summer. It is entirely possible that I missed it as I attempted to wade through thousands of posts in six threads, but has the position of the PF's armrest been discussed?

TTex600;

There has been on and off discussion regarding possible reasons for the PF adopting the NU inputs with the Sidestick.

Takata posted full details and description of the SS in AF447 Thread No.5 - post# 1818, and using the following Google search format will reveal a number of other pages of interest:-

af447 pf arm rest site : http://www.pprune.org/tech-log/

Re: armrest. Whether a deviation from ergonomic neutral, or not, something chronic was at play, and makes the PF look stubborn to climb, when he very well may have been stubbornly seeking level. The a/c demonstrates a chronic Nose drop and right roll throughout the descent in final flight path. Each of his (Initial) corrections could easily have been misinterpreted by him, and been incremental climb orders instead of 'returns to level' (or to 3 degrees PU). His screen was not recorded.

AoA meters ? But isn't the opinion that just holding pitch and power would have been sufficient to avoid the stall ? Once stalled, with the stall warner sounding, altitude bleeding off and nose high, it is distinctly unclear what another instrument would have added. Admittedly I can see how it may help recovery to optimise pull out and avoid secondary stall but given that the automatics were actually announcing 'stall' it isn't clear that having a readout which could be interpreted as showing a stalled condition would have helped given all the clues available anyway.

Isn't the more immediate problem to understand why, with all the 'help', the stall condition wasn't recognised and if the 'help' wasn't actually a hindrance.

There are several reasons why the STALL went unnoticed. And once entered, why it was not recognized. There is something missing.......

Offhand, I'd guess attitude was unavailable. One can continue to ask why no one sussed attitude; the only likely explanation is that attitude was not reporting. OR, it was reading NOSE DOWN, somewhere. If you believe it was available, and accurate, that you would even consider flying again qualifies as an insane thought. One cannot believe these three were that thick, and no others are.

think about it.

Well, I am SLf. I keep posting the same thing about stall and AoA every 3 months or so. Can't we move on or remain silent in the absence of new information ?

It is or it should, but as it is obvious it did not work out well in this case and the question remains why? What is necessary, that the simple application of a procedure "pitch and power" works out as desired?

- recognize the event UAS
- be in a steady state of flight when UAS occurs
- recall the correct procedure for UAS
- apply the procedure UAS in a correct way
- monitor the outcome of the procedure
- recognize the point when speeds are valid again
- go back to normal

Imho with a AOA gauge it gets a lot simpler, as soon as something looks wrong with the performance (speed, pitch, power) the AOA value becomes the primary performance value. Keep it in the desired range by increasing or decreasing pitch, bank, loadfactor, speed, power, altitude. Use whatever means are necessary and applicable to the situation to maintain this value and you are on the safe side.

Using the AOA value is using a dynamic value with lot of options instead of limiting the options to two parameters, namly pitch and power without possibility to verify the correct application (with the doubt of overspeeding or getting too slow).

Using the AOA value there is no necessity to start out with a stable situation (straight and level unaccelerated flight), the usage can start in any flight situation.

That does not only apply for UAS, but also for other situations (engine out comes to my mind) where performance gets critical.


Sorry, i´m having a hard time to understand that one in context with with your previous post. It is true, we are in a hamster wheel of questions surfacing again and again, but it also makes no sense to let wrong statements or wrong assumptions unanswered.

You yourself as a paying customer in aircraft (i find the term SLF neither appropriate nor respectful) obviously cannot understand the difference it will make for an aircrew in the pre-STALL timeframe, where an AOA gauge can give already detailed information about the performance state of the airframe and hence can prevent the triggering of the stall warning beforehand.

There are some pilots out there who had an AOA gauge available like myself in the military and who know how easy to use and how helpful it can be. Others never saw one and never used one (except some glider expierienced pilots, the tread on the side of the canopy is nothing else then an AOA indicator in the purest form), but will have no problem at all to understand its function and implement it in the crosscheck.

But now i go "silent" to lurking and reading like i did the past 5 weeks (as there was no new information) and thus following your suggestion.

franzl

PF Armrest
 

Hi mm43,
That link produces "Page not found" for me, unfortunately.

TTex600 is right to raise the armrest position as a possible issue. Despite much discussion of the sidestick on these many AF447 threads, I don't remember any of us describing or discussing the armrest specifically. In my own several attempts to explain good and bad techniques for using the sidestick – and to discuss the possible difficulty of coping with the AF447 combination of pitch-alternate with roll-direct – I neglected to mention its importance.

To use the sidestick accurately, it's vital to be able to relax the shoulder and arm completely. This is particularly important in turbulence. The only movements required are of the wrist, hand, and fingers/thumb, which may involve rotations of the forearm. Although these may produce limited sideways movements of the forearm, these are only by reaction. The forearm should remain relaxed. The elbow must remain stationary, and be supported at a height comfortable for the pilot.

The pilot's seat position (fore and aft; up and down) is itself dictated by the requirement for the eyes to be in the correct position as indicated by his lining up two small balls mounted on the centre windshield pillar. It follows that the armrest has to cater for a wide variety of elbow positions. It achieves this by being adjustable for height, which means it also needs adjustment for rake. Both axes have a graduated scale. On conversion to type, pilots find the most effective position and memorise the alpha-numeric reading, which can be quickly achieved when entering the seat. As a line-checker, I was not always convinced that a pilot had found his/her optimum position.

It follows that, during a seat reshuffle, each pilot needs to reposition his new armrest. Sadly, we shall never know if armrest position might have been a factor in the mishandling of pitch in AF447. If memory serves, however, the PF had not changed seats during the flight.

and AF471 (Caracas-Paris) had also an armrest problem? While never in the history of a sidestick aircraft there was an incident with armrests as a contributing factor? Very unlikely, me thinks.

I think it is interesting that there is a record of 5k fpm DOWN (VS/sel) in the FDR up til a/p release, then the problem went to manual, and the pilot selected repeatedly NOSE UP.

Approaching the chronic aft stick as an ergonomic issue is interesting. But, yes, the PF had been in his seat since T/O, so the likelihood that a cause for chronic incorrect bias (if there was one) in PITCH may have been elsewhere?

I think it is unfounded, to dismiss a thus far unknown potential for chronic a/c ND that caused PF (and PNF, had he the stick control) to desire constant NU, as correction. Remember the PNF was seeing the results of PF's actions, not the SS position that initiated such result. So PNF would not necessarily have seen the attitude of the NOSE as different from the results of control from the right side. Without seeing the RSS, PNF can not know cause/result separately. He sees results. The PF's panel was not recorded.

Chris, I figured the URL out. The software is fooling us into just clicking on the hot link. We need to copy the entire line into a search engine. Such as " armrest af447 site:http://www.pprune.org/tech-log/".


It will likely never be known, but considering the wild, full deflection SS movements the PF made throughout the event; I have to wonder if the armrest was UP. Not whether is was properly adjusted, but if it was even in the down (required for flying) position. The PF could easily have raised it after reaching cruise alt for any one of many reasons (dropped a pencil, dropped a chart, dropped a M&M) and not lowered it again.

For the flight sim'rs and SLF out there, the SS side armrest must be down and properly adjusted in order to hand fly the aircraft. This is to isolate the pilots hand and wrist (which manipulate the SS) from the rest of the pilots body. With the forearm resting on a properly positioned armrest, the pilot can manipulate the SS with precision. When one attempts to hand fly with the armrest up, PIO's are hard to prevent because flight loads move the pilots body which moves the arm which moves the hand which moves the stick. ;) In my case, it was at rotation in smooth air on a VMC day and I was able to drop the arm rest before the gear came up, but SS control was compromised for sure.

FWIW, I will defend the pilots until it's obvious to ME that they were the cause, but even I have a hard time defending the continuous full scale deflections of the SS. Every Airbus SS pilot with any hand flying experience understands that the you don't "stir the pot". "Stir the pot" = slang term for excessive and inappropriate full scale deflections of the SS.

edited to fix typographical errors and add caveat

Caveat: I'm speaking of the 320 when I speak of the armrest. I can only assume that the 330 has the same set up.

But once the aircraft was stalled and 'mushing', the controls wouldn't be having the effect the PF expected so couldn't he quickly move into a more extreme, try anything, this isn't working type behaviour ? In these circumstances a bit harsh to think this was all he ever was capable of in any circumstances ?

@TTex600

I think the armrest angle question was discussed at length in one of the Tech Log threads. No conclusions could be drawn, obviously - but if I recall correctly not a lot of the 'bus pilots thought it was likely given the extremity of the deflections made.

If memory serves, PF replaced the Captain in the Left Seat (LS) after top of climb. PF would expect to relinquish the LS to the Captain later in the cruise, well before top of descent. PF would not have expected having to hand-fly the aircraft. Perhaps, as a matter of courtesy, he left the Captain's armrest settings in place. It would be interesting to compare the physical dimensions of those two pilots.

Nope, the PF was in the RH seat, as indicated by the CVR and the DFDR traces - and had occupied that position since pushback. The PNF replaced the Captain in the LH seat, having taken his rest at the start of the flight.

even if you have your armset wrong, over even up, or superglued, or frozen, or full of M&Ms, you would as a normal pilot never pull while your aircraft is 15° attitude nose up and you have stall warning (might be not be audible because of the big fuss in the cockpit).

These guys pulled because they wanted to pull. There is no remark on the tape that there was something wrong with pulling (only later when they were completly stalled). The same seams to be on the other AF incident I was referring to (AF471).

That's a strong indication to me that they thought pulling is good.

There is no reason I can see to claim the PF was not attempting 3-5 degrees UP. Is there a claim he left the thrust alone by mistake? Was it at ~82?

If there was turbulence, perhaps some leeway for the PF not identifying the intital 1.65 g? How is he to sum the g to frame his attitude, and what did he see? Is he Mr. Inertial?

Seen it before; Instant Conclusion, then the search for evidence to sustain. Sometimes it is useful, but only when one understands it is for purposes other than conclusive report. EG: Discussion. Much remains of the evidence that is hidden.

The JUDGE is not allowed to see it? RED FLAG......

Hi,

Neptunus Rex
:confused:
If someone has such thoughts .. it must never sit in a cockpit as pilot .. never

Do you know that? Do you know none of the pilots had maneuvered her prior to a/p loss? Is there an unbroken stream of autoflight? You know this?

As an Airbus (320 series) pilot, I say that flying with no armrest might be one of the only reasons one could explain such clumsy SS treatment.

Any Bus pilot with any degree of hand flying experience understands that rapid SS movement, even to full deflection, produces little in the way of aircraft movement. I have a friend who's flown the Bus on m/x flights who tells me that he can move the stick, rapidly from corner to corner, side to side, on approach and that the aircraft doesn't roll or pitch at all. I've never flown it without pax, so I can't verify, but I can verify that full aileron in a gusty crosswind accomplishes no more than just leaving the stick neutral.

that's absolutly correct, and I'm often amazed of how some of my collegues "stir the pot", i.e. fiddle around with the stick without any (great) effect. Real Airbus flying is: give an input, look what it does, give the next input (if needed).

But I don't see any stirring with AF447. He pulled fully and hard. If he would have "stirred the pot", you could see a zick-zack line that changes direction within split second, similar to the "zipper" some see in the vertical speed artefact. If he would have "stirred the pot", he would have given down inputs at least 50% of the time. But he didn't he pulled most of the time. Because he wanted to pull.