BOAC
I got that "stall warning triggered by low air-speed indication from -
Initial Air France Flight 447 Black Box Info Raises More Questions Than Answers: Pilot | NYCAviation

I was a bit surprised because I had thought stall was purely a function of AOA, and no doubt air density as well. But I can see that air speed might come into it to, but I would have to have the mechanism of that explained to me

Before adding AoA crosscheck to the "pilots tool kit" it needs to be added to a display somewhere a pilot can access it.

Unless I misread something UAS blanks out the one "proxy" AoA display available to the pilots.

Dozy:
The human factors element of this mishap seems an apt parallel to what AF 447's PF might have been experiencing.

@Murphy:

Yes, of course, sorry I left that out. As I noted earlier, choosing what piece of "real estate" in the display area would house an AoA gage requires careful consideration, and some ergonomic and "scan efficiency" study before a decision is taken.

From the report:

The question is has the system rejected a/s (60 kt) by this point? If so, why is it still being honored by AoA logic?

"When they are below 30 kt, the speed values themselves are considered invalid."

Considered invalid by which sub-system? AoA logic? BEA doesn't say.

Welsh Wingman
 

"possible weather deviation limitations for this plane on this particular route (unless the flight crew want to land in Bordeaux and endear themselves to flight ops...)"

Weather deviations should cost peanuts in fuel quantity (ask your math teacher) so can we put this one to bed please...

LW_50 :

Many thanks. I've been saying this since the tail-end of the previous thread, when interest picked up over the BEA's "note", but very few people seem to have noticed. In addition, I was also clued in to the BEA's report on the Moscow A310 incident, which referred back to data that says that without proper training, 80% of pilots instinctively pull back on the yoke or stick when confronted with an unexpected stall warning.

What a load of twaddle!

This persistent habit of inventing stuff out of the blue, and presenting it as it is somehow fact, while ignoring or distorting what is known fact, to allude or insinuate that the aircraft suffers from some unspecified flaw and that the BEA is in cahoots with the establishment to somehow pin it on the crew is really getting old.

Give it a rest, for everybody's sake.

Doze

"Bear, you've got to be careful how you phrase this stuff. A newcomer to this thread would think that what you're presenting is a known fact rather than a theory posited by a single poster that is based on no evidence whatsoever."

There is no evidence as yet. Evidence is entered under penalty of perjury, and as "held and factual".

Here, we have ACARS, some few motes of "data", and miles of "O'Briennese" (no offense, mate). If a/p dropped, any flight vectors that had been under control, would have reverted to untrimmed, or been patent under a/p commands. Thus, "Trending". Not obvious? Fine.

Any one who is misled by reading any of this, needs a therapy. Informed opinion, at best. You give at least this old pilot too much credit.

happy day, Sir.
bear

ec, one outsider........


"This persistent habit of inventing stuff out of the blue, and presenting it as it is somehow fact, while ignoring or distorting what is known fact*, to allude or insinuate that the aircraft suffers from some unspecified flaw and that the BEA is in cahoots with the establishment to somehow pin it on the crew is really getting old."

*"known fact"........Sorry?

jcgeant
"Sorry if I'm dumb .. but .. someone can make a technical comment (explanation)
about the text in bold .. ?

From the 27 May BEA report:
Quote: Around fifteen seconds later, the speed displayed on the ISIS increased sharply towards 185 kt;
it was then consistent with the other recorded speed. The PF continued to make nose-up inputs. The airplane’s altitude reached its maximum of about 38,000 ft, its pitch attitude and angle of attack being 16 degrees."

That is puzzling. If the wing chord angle is zero it makes sense - the plane was moving horizontally through still air. I can't see how it can ever happen if the chord angle is not zero with a horizontally moving plane (which is implied by "reached maximum altitude").

Bear, don't be obtuse. I meant evidence in general terms (as in a fact or collection of facts), not legal.

[EDIT : I note that you're flying your true colours more obviously on the sister thread in R&N:

]

its geometrie: in still air if the flightpath = chord angle (mayby + 3 deg....) than pitch attitude= AoA

@Hazelnuts39: I assume, you know the A330 and its related non-normal procedures: It says in QRH 2.21 "Respect Stall warning and disregard "Risk of undue stall warning" status message". In so far, you would be right. However, if you follow the whole procedure, crosscheck your Attitude/power-setting according to the table provided in 2.23, crosscheck with the GPS-derived speed and get the stall-warning nevertheless (BTW: with changing pitch angles) you consider the stall-warning invalid. Even the TRE explained that the simulator does not behave according to the manuals.

All I wanted to hint to is that sometimes the (aural) warnings created by the aircraft are adding up to the pressure you already have and should be able to be cancelled (e.g. by use of the emergency cancel-button, which did not work in this case...)

Doze.

Quote:

I know that the increased presence of computers in the flight deck is an emotive issue for pilots, but I can assure you that from the perspective of this software engineer, and, I'd be willing to wager even more - every software engineer who worked on the FBW aircraft that are flying today - we are on your side.

Amen. And thank you...

Airbus control logic and alternate mode control
 

One problem that I see here with the Airbus approach to degraded mode operation is that the aircraft can be put into a flight mode that is not natural. In normal law the aircraft flies to a point in space where the pilot tells it to. In Direct Law the airplane is the same as a conventional aircraft, but in Alternate law the aircraft doesn’t have enough information to control the aircraft in a stable manner and it relies on the pilot for pitch stability. That is, when flying in normal mode the aircraft is controlled to point in the sky, and the autopilot system controls flight path by adjusting power and trim to match the control inputs of the pilot and the speed selected.

That’s all fine and good until something goes wrong with the inputs. When the control system goes into Alternate Law things get much more difficult. What I understand happens in that instance is that the pilot becomes the one who is controlling the aircraft in pitch and, the system is not stable in pitch. With a conventional trim system you set the trim angle and if you release stick pressure the aircraft will return to the speed required to match that pitch trim setting. This is the definition of a fundamentally speed stable system, remove the inputs and the aircraft goes back to a trimmed airspeed. With such a system, if the pilot made an incorrect input or responded to turbulence, all you have to do is relax stick force and the aircraft returns to a stable condition. With the Airbus the pilot must maintain pitch control with reference to the HSI and the trim will move around. Some of the Airbus pilots here say that isn’t excessively difficult, but it is surely a much higher workload to constantly keep the required pitch in the HSI while you are trying to fly through turbulence.

The Airbus system, in Alternate Law, will adjust the THS to “chase” the pilot inputs by changing the trim position. This is fundamentally unstable and consequently, unless you make absolutely no forward and aft inputs, the aircraft will climb or descend and the speed will increase and decrease. If the inputs are large enough and long enough, the trim will change and the aircraft will remain in a climbing or descending mode until the pilot applies input. It seems to me that Alternate Law without airspeed is really kind of a bastard system in that the control really doesn’t have enough information to fly the aircraft, so the pilot has to step in and provide stability.

In my opinion, the proper control system response to loss of airspeed information would be for the autotrim to be disabled at the same time as the autothrottle and set to the same condition as it was just before the autopilot dropped out. So I guess that in the case of loss of speed sensors I am thinking the system would be much better off in direct law.

Then the aircraft will be remain at a known pitch condition and therefore be speed stable. You don’t know what the speed is, but it was fine where it was and it isn’t going to change. The pilot then only has to control altitude with the throttle and keep the wings level and he is in a much safer place. If the pilot decides he wants more or less speed he can adjust the trim wheel to get it.

Somebody please correct me if I am wrong.

@ engin-eer:

The autotrim in Alt2 will only set the THS to "chasing" inputs if that input is held past the limits of elevator authority, otherwise it simply stays in the last position set when the FMC/autopilot kicked out. Letting the stick go in that situation will result in what you describe - i.e. the aircraft trimming itself back to it's stable position.

The only reason the autotrim moves to match pilot input is to give maximum control authority to a pilot in a significant upset without having to have his or her hands on both the sidestick and the trim-wheel. It follows that to have autotrim responding to sidestick inputs requires an emphatic stick movement held for some time.

Yeah, I got that; but the flight path is horizontal at the top of the path.
You could only get pitch = AoA in horizontal flight if there is an upward wind.
Maybe there was and that's what helped to get the plane up there in the first place.

Old China
 

Cross purposes, I think.

"
Welsh Wingman
"possible weather deviation limitations for this plane on this particular route (unless the flight crew want to land in Bordeaux and endear themselves to flight ops...)"

Weather deviations should cost peanuts in fuel quantity (ask your math teacher) so can we put this one to bed please... "

I didn't mean the particular de minimis deviation cost in this instance, but rather in the wider systemic issues that I was discussing. I was under the impression that this plane was close to MTOW at GIG and this route is close to its maximum range (never flown an A330-200 or any other Airbus before retiring in 1992, so not claiming any expertise), so merely hypothesising whether the BEA are hypothesising about the wisdom/relevance (if any) of even sending pilots into the ITCZ in the knowledge that a "major" weather diversion could mean a refuelling stop short of destination. I recall an old SAA friend telling me a few years back that some SAA pilots were frustrated at having to fly the A340-600 on the Joburg/New York route, when a longer range A340-500 would be perfect for the route if SAA had properly invested in the right equipment for the route.

Feel free to shoot this down, now that I have made clear. Not arguing that this is likely to be one of the major factors in the hull loss, but wouldn't yet discount it altogether from the overall mix in the context of not always giving equatorial CBs the healthy respect they deserve (unless I am wrong about my understanding of the range of the A330-200, which I accept that I might well be).

Not wanting to send anyone off on a wild goose chase.

I was thinking more about my old history teacher rather than my math teacher - divide-up the understanding of historical events into long-term causes (aviation industry focus?), short-term causes (aircraft manufacturer and/or aircraft operator focus?) and triggers (aircraft operator SOPs and flight deck focus?)....

Hi,

Well I asked a explanation from pilots as I was puzzled by "Level 380" ... because this other A330 case (AF445) same sector .......

They were descending .. or they were stalling ?
Note that all recorded datas of this flight were never examined by any regulator or the BEA (AF was not able to produce those datas)

@ engine-eer, I agree. In direct law I'd have a conventional Boeing with natural speed stability - perfect.

@ dozywannabe. Alt Law in pitch = "In flight, the alternate law pitch mode follows a load-factor demand law much as the normal law pitch mode does, but it has less built-in protection".
It doesn't sound much like direct law to me.

Dozy:
I understand the advantages of autotrim, but I have reservations about the reason autotrim matches pilot input as explained here.
Pilots have two hands. There are two pilots on the flight deck. Four hands. Autotrim available in an upset seem to me a very handy tool, particularly as the pilot (as I understand it) can override it via side stick commands.

Would it make sense that if one is in a Law or condition that calls for manual trim, that procedures would call for the PNF to be on the throttles as required/requested by the PF? That leaves PF with one hand on side stick and one hand on trim wheel.

I am in a CRM zone here.

That strikes me as how to avoid worrying about a two handed pilot having a three handed job, which you seem to allude to above.

Does the approach to this known possible condition vary between companies, or is there a generally agreed standard on how to deal with this in Direct law, upset or no?

Ok, I get it, he told the airplane to climb and it did. In this particular case Direct law would have been better than Alternate law since it wouldn't have trimmed him nose up, but if he hadn't held aft stick to the moon he wouldn't have been changing the trim either...

Somebody once said "it's hard to make anything foolproof because fools are so ingenious"......

@rudderrudderrat

I didn't say it was. You need to check the difference between Alt 1 and Alt 2. In this case they had Alt 2 + airspeed instrument failure, which pretty much nullifies stall protection. In any Alternate Law mode, the protections can be overridden if the stick input is firm enough and held for long enough. In short - in anything other than Normal Law, do not hold the stick back expecting the protections to have your back.

@engine-eer

Hi - cheers, but that's not what I'm saying. I don't know what he did (outside of the points in time mentioned in the BEA "note"), and neither does anyone else on this thread that I know of. I'm deliberately keeping any speculation to myself as much as I am able, because then I'd be no more useful to the thread than people who are saying that it must have been the pilots, the computer or even the weather. All I am saying is that the systems behave in such-and-such a manner if they are operating properly (i.e. as specified and designed).

I am a bit puzzled about the concern about AoA indication. I presume maintaining what you guys call 'pitch and power' would probably have saved the day. If the climb was intentional then that's a step in the dark and the inability to adjust thrust to keep air speed up was the issue (not easy without airspeed reports). If the climb wasn't intentional but went unnoticed, seems unlikely that additional information on AoA would have been recognised. In the stall, if it wasn't recognised as a stal,l then what chance an AoA readout would have been noticed/acted upon. If it was recognised as a stall the -40 degree or so readout would have been irrelevant.

On another point, if the systems had some remaining degree of automation, given a pilot input of nose down, would it be so difficult for the system to recognise the intent, recognise the THS setting, and help him out by unfixing and adjusting it ? Seems a bit perverse to freeze the setting and not revisit it in the face of sustained pilot command.

Either the information provided by the BEA is inaccurate, or I have forgotten my trig, or there is a persistent misunderstanding on this thread about the attitude and flight path. The BEA report says pitch is 16 degrees nose up, 35-40 degrees angle of attack, and 10,000 fpm down. To get the flight path angle, you subtract the angle of attack from the positive pitch angle, giving around -20 to -25 degrees. The true airspeed is 10,000 fpm, or around 100 knots, divided by the sine of the angle, which gives around 250 knots. The airplane was in a stalled, mushing descent, not a vertical plunge or falling-leaf type of maneuver. At least this is what the BEA numbers require.

It appears to me -- with zero experience of flying this type of aircraft -- that recovery would not have required a great deal of height, once the wing was unstalled by trimming the THS and getting the pitch angle into the vicinity of the flight path angle, ie 20 degrees or so below the horizon.

jcgeant, see:
http://www.pprune.org/rumours-news/3...yday-call.html

see particularly mm43's posts in that thread

Mr Optimistic, I respectfully disagree. Once you've used, or gotten used to, AoA as a gage, or been trained how to use it, it becomes another instrument you can go to for information ... like your engine instruments. Since the information it holds, stall or not stall, is somewhat important to flying ... "AoA high, lower the nose" isn't too hard of an association for a pilot to make. Hearing a stall warning tone is not the same as seeing your AoA. (For that matter, rudder shakers or stick shakers are another place where a sense other than hearing is used to cue a pilot that he is stalled or nearly stalled ... )
Presuming AoA is measured in degrees, not units, I follow what you are saying. The velocity seems to have been enough to return to flying if the nose could have been aligned with the flight path angle (presuming enough elevator/THS authority is available to get the nose pitching down).

Is that part of the point you are making?

Yup. (+ six additional characters)

No THS Position Indicator??
 

Do I understand correctly that there is no THS position indicator?

If not, we have a situation where the "system" trimmed the aircraft into a stall, something we learned early on not to do, and yet the pilots had no indication of that.

Sir, it seems that there is a THS indicator (visual), if one looks down at the trim wheels next to the throttle quadrants. My diagram (old) does not have enough fidelity to tell me how the wheels are marked.

http://avionique.free.fr/IMG/jpg/Cockpit_A330.jpg

That picture (if it is a faithful rendering of an A330 cockpit) shows me what looks to be an index, and large white blob (??) and what may be a scale in green.

Also recall in numerous threads comments about AB pilots needing to set the trim wheel to a particular value for takeoff (probably varies with gross weight??) So they'd need to be able to read the trim wheel's position to confirm their nose trim is at the value they calculated and require for takeoff.

lonewolf: thanks. I am not a pilot so I am in no position to argue ! But if pitch, power and altitude rate were apparent and not taken due accord of.....?

Stall warning
 

The maximum angle of attack (i.e. where the plane stalls) must depend on a number of factors - pressure altitude, airspeed, load factor, aircraft weight, config. Lots of points of failure for a measurement system.

I'm not sure if actually displaying a raw AoA it to the pilot would be so useful - otherwise every C152 would have such an instrument. How would the pilot know what is the current alpha-max?

The Airbus FCOM states that in case of a simultaneous OVERSPEED and STALL WARNING the stall is to be actioned - failsafe: if a pitot fault gives us an excessive speed the alpha-max will decrease, so if we follow the stall warning we basically end up flying a lower AoA then would otherwise be possible. But the 'under-speed' scenario does not seem to be addressed too well.

Now the types I am more familiar with - the C152 and DA20 - have a stall warning system which is basically an orifice on the wing connected by means of a tube to a little buzzer - it detects when the air pressure on the wing suggests an oncoming stall. This system is far superior to what the A330 has in one respect - it is completely independent of the airspeed system.

Now - why can't the Airbus/Boeing engineers devise such a system for the big jets? I do not suggests orifices, tubes and buzzers - rather - pressure transducers (or perhaps strain gauges?), a data transmission network and a clever computer to work out what is actually happening with the wing. A system which tells us if the wing is generating lift or not, how much lift, which part of the wing is stalled.

A330 Manual Pitch Trim
 

The manual pitch trim is mechanically connected to the THS. That THS is pretty big.

Just out of curiosity - approx. how many revolutions of the manual pitch trim would it take to bring the THS from the max nose-up angle (12 degrees) to an angle where elevator authority is sufficient to command pitch down (8 degrees I recall). What about 0 degrees?

What force is required to move the pitch trim wheel? It says: "Mechanical connection", but is there some kind of power assist?

If the THS jammed, would the wheel also jam? Can icing cause the THS to jam?

What is the relation between manual trim and autotrim? If I manually trim does the autotrim disengage? When does it re-engage? What if autotrim is commanding nose up and I trim nose down - who has the last word?

At the point of stall Flight law Auto- Trim was active for all that we know.
The only point where it might have reverted to manual was when Alpha = 30° was exceeded at around 2:11:35.
If the BEA 'leak' is correct Auto Trim was even active throughout the whole descent.

Golf Sierra

The Navy aircraft I referred to earlier get around that with units ... not sure if that is applicable for transport aircraft.

Good points, also the understanding that AoA and stall AoA are not fixed values, but vary as you point out.

I am not sure I understand this.
By "otherwise possible" do you mean "allowed by the protections" or physically possible? Is alpha max a computed value, or is it the raw AoA measured by the AoA probe?

My point is, if you have a false airspeed indication, the AoA probe/detector doesn't give a rat's behind. It is measuring something other than dynamic pressure, and if I may say, it is measuring something simpler: angle of airflow relative to the aircraft. (Or should I say airfoil?) Of course, if the AoA probe gets stuck/frozen rendered moot someway, then ... well, that's another dog's breakfast served to the flight deck crew. (Gee, it goes back to pitch and power again. :cool: )

EDIT: yes, there are two different hydraulic boost systems that help the mechanical linkage move the THS.

This picture purports to show A330 AoA probes, which look a lot like AoA probes I've seen on other aircraft. They do not measure what the pitot probes measure. What appears to happen inside the computer in the A330 is that airspeed input is compared to AoA probe reading is compared to other data and, once the calculations are done at the speed of electricity, visual cues are presented on the strips (vertical) next to the pilots attitude indicator on his flying display.

(VLS/V alpah prot V alpha Max)

Golf-Sierra -- you may find this a good place to start:

http://www.smartcockpit.com/data/pdf...t_Controls.pdf

CogSim
Stall is due to too an AoA which is outside the boundaries, but stall warnings are in addition due to your AoA being in danger of going outside the boundaries.
If your horizontal velocity is very low you have little lift (even if not stalled) and your downward component of velocity will increase so much that the AoA will go into the stall region.

Loanwolf your #1149
 

There was a reference [url] from JT to an earlier thread post from PJ2 detailing the 330 trim wheels on the pedestal the integrated read out of THS trim angle along with the follow up information available on the centre display, can't remember the actual display name 100% so will not say.
all the best

t54, that was me, not cog sim, on that AoA tidbit.

FWIW, I see no reason to divorce stall warning from your AoA, but if you are working in raw angles versus units, I can see the issue becoming more complicated as the critical AoA is related to so many other factors. Thanks.

One way or another, be it at or beyond stall value, or "close to stall value, do something mate!" a suitable warning ought to be made.

Uh, not quite, I am not sure I understand you there. You can go slow (low horizontal velocity) and level if you are not stalled, and go merrily without falling, or stalling.

That said, once you begin to descend, the second part of your sentence begins to make sense.

If what you are saying is that you are already falling, and have a low horizontal vector, then sure, the vector sums (depending upon rate of descent) change your AoA, and not necessarily for the better.

What has this to do with AoA indication or warning? AoA probe will measure how the air is flowing. (Direction)

True.

I think we all have to be very careful to differentiate between the terms "Stall Warning", "Stall Protection" and "Stall".

The first can work to an extent without airspeed data but the second absolutely cannot. So if you get into the first or third in Alternate Law with no speed indication you are going to have to think very hard and work very methodically as quickly as you can.

Mr Optimistic:
While true that pitch and power would likely have worked something else -did- go wrong and the pilots ended up in a confusing situation.

At that poin a raw AoA indicator would have likely helped them to sort out
the counterintuitive return of stall warning as they were applying nose down.

In fact the stall warning in this case was a tragically misleading on/off AoA indicator that had 3 states:

Normal : silent
close to stall or stalled : STALL STALL STALL
severe stall : silent

Severe stall results in <60kt sensed speed and inhibition of warning.

This is not related to pitot ice/failure, it is a combination of true speed plus AoA effects on pitot as others have pointed out.

Had the raw data been avaiable they likely would have seen the AoA improving as the warning returned and pressed on with the correct recovery.