AoA Protection
 

@a33z
Very good description of the "intent" of the AoA protection. It's what "protection" should be.

My question has to do with at least one reversion mode in which it appears the system establishes an AoA other than max and more than what is required for level flight. In other words, the gee bias is reduced as far as trimming to the normal pitch/roll-corrected one gee.

This action would appear to move the THS trim posiiton, with or without pilot holding other than neutral stick. So I question, "why not maintain the basic gee command and only use AoA only as a limit?"

Secondly, there are some very good reasons to use less than the AoA required for level flight - like recovering from a stall!

BTW, ol' Gums did not call the system "ridiculous", that was another contributor. I still feel the reversion laws could be streamlined and some work could be done with respect to "standby gains" when air data is unreliable.

@A33Zab re Probe heating
 

A33Zab re #551
Described monitoring is task of the installed Probe Heat Computer (3x)

Can you supply a reference to the functionality of the PHCs; and the warning messages they can generate?

My searches have had very limited success so far. However they do not point to any PHC functionality other than ensuring the pitots have voltage applied when appropriate and are drawing reasonable current (i.e. the element is not burning/burned out?).

So far I've not seen any indication that the PHC does anything other than control the normal operation of pitot heating, and monitor for faults.

My suggestion was to infer the presence of high-altitude icing conditions from the pitot's electrical behaviour; which in turn reflects the thermal load on the pitot. I'm delighted if somebody is already doing this, but have seen no indication of it in any of the reports of high-altitude pitot icing I've read.

Would you know if that technical characteristic was mentioned in the manuals for the 320, or was that information directly produced by Airbus for the Perpignan report ?

• inadequate maintenance ?
• lack of maintenance ?
• degradation due to aging ?
• altered manufacturing process ?
• ... ?

The reported problems on the GOODRICH are very few.

HN39;
Unless I have misunderstood your comments I'm not sure where or why my views weren't clear on pitch and power. I think we agree completely on this, so permit me to clarify, and if I have misunderstood your thoughts, please help me out, ;)

Specifically on your point regarding 'the better strategy', contrary to trying to control either speed or altitude within tight limits and short time frames, letting airspeed and altitude "float" in response to turbulence sufficiently strong to displace the aircraft has been an SOP and a trained response for a long time.

From the FCOM, the Turbulence Penetration SOP, in part, reads:

Set the thrust to give the recommended speed

(Refer to QRH). This thrust setting aims to obtain, in stabilized conditions, the speed for turbulence penetration.

Change thrust only in case of an extreme variation in airspeed, and do not chase your Mach or airspeed.

A transient increase is preferable to a loss of speed that decreases buffet margins and is difficult to recover.

InfrquentFlyer 789, "Clearly if above 30kft, we shouldn't be needing a GPWS escape, so maybe disable it. Then what happens if the plane has dodgy altitude info ? Ah, but what's the probability of needing a GPWS escape and at the same time the altitude being screwed ? "

Um, er, ah, please explain how a modern plane could experience dodgy altitude readings over water in the middle of the Atlantic. And do the extent of small thousands of feet even over any place on the face of the Earth. (Small 10s of feet if good DEM data is used along with GPS.)

Now, if you need barometric altitude for aerodynamics purposes I can understand it. However, for the simple task of noticing you are falling very rapidly GPS should be more than adequate for the job as should the ring laser gyros and accelerometers on the plane.

For a couple minutes GPS/Inertial data should be perfectly a adequante substitute for air speed once translated.

jcjeant, "BTW .. all Airbus accidents so far are due to pilots errors."

And it seems nobody important enough is asking the obvious next question, "Why?"

Airbus can make a good plane but they can't make a good pilot

Peter H

There is a bit more info in the design specification of Airbus pitot tubes systems in the ATSB Guam report AO-2009-065 (involving the preferred Goodrich probe).

but again it just mentions monitoring current against thresholds. To justify the term Probe Heat Computer I trust it measures true AC (115V 400Hz?) rms power using V*I*cos(phi), but that used to be done in simple hardware devices. There should be a measurable change in heater resistance with heater coil temperature, but guess not simple to compensate for 'wind chill effect' of different humidity air streams - its also possible external ice build up creates a thermal insulating layer. If you just want to detect icing conditions use an ice sensor- melting point close to 0 deg C despite air mass at -40C at cruise altitude?

Then as previously mentioned they have been fishing a lot of dead Airbus trained pilots out of the Med and Atlantic that all must have errored since the Airbus is not at fault.

PJ2

Very well explained discussion re airpeed inop flight.

Do they practise this in the simulator? Do they practise flight in all the various laws?

galaxy flyer, :sad: if it is so hard to fly AirBus fly by wire because pilots learn to fly as God and the Wright Brothers intended then how on Earth do those who manage to transition to F18s, F16s, and other fine fly by wire aircraft able to manage the feat and do it so well they can out fly any other aircraft in the world?

I realize these young men are minor gods. But that should not stop "mere morals" who might find themselves at the stick of a modern FBW aircraft from doing a better job than we see with AF447. And, as I harp on excessively, "WHY do we see this, repeatedly?" Are the FBW rules wrong at some point? Is the training wrong? (That is a VERY real probability considering that AB and AF decided to change the stall training for all pilots.) Yatta and more yatta.

I would like to learn why the pilots screwed up flying the plane when it was handed to them in pitch darkness with no speed indications and zero visibility. They committed some blunders, TWO OF THEM committed or permitted some blunders (without screaming at the top of their lungs or physically assaulting each other). Why did this happen? And blaming the FBW and side stick is not going to cut it with me. It was bad decisions. Where did the FBW do something the pilots did not tell it to do? Why did they persist until at least some here (I'm not sure) suspect the pilots saw light glinting off the ocean and BOTH made ND inputs.

edit: Oops - not F4, yes F18 (and many others.)

{o.o}

(mumble) 1g (mumble) (mumble)
 

Grity, Gums, et al.

I presume you are using "1g" as a shorthand for 1g normal to the surface of the Earth, right?

If not I can see a pilot getting into trouble with an indicated 1g acceleration vector that points off somewhere other than straight down in the direction of the Earth's gravity. Worst case, of course, would be 1g straight down regardless of the heading of the aircraft.

My fool head sees 1g with no direction vector and simply cringes.

RR_NDB, for all your asking for reasons for the pitot redundancy you surely don't think that the failure was in the pitots rather than the obvious common mode failure of the air itself, do you?

If this icing can affect a Thales probe it can also affect any other probe if conditions are "just right." So simply changing probes is more like sweeping the problem under the rug than solving it.

You're on about just one of the contributing factors that led up to the AP/AT disconnect. Yes, any one of those factors changed could have changed the outcome here. It won't change the outcome in some other cases.

Some of the things I can see for improvement (with my personal bias to "communicate") include:

• communicate not shrug when DAKAR does not answer,
• deviate around storms, know your radar (radar training),
• pseudo-airspeed (GPS derived for as long as altitude and ground speed remain "sane"),
• a variety of probes (sadly not possible as only two are qualified),
• modified pilot training (What do I do when a stall warning appears? What do I do on AP/AT disconnect?),
• reduce cockpit information overload in crisis (obviously conflicting warnings MUST be solved),
• and more.

I am sure people here can see more. And some of the above may involve solutions too expensive for an accident that for all its horrible nature is way down in the statistical noise for historical levels of air safety.

(The AirBus is a remarkably safe airplane already, possibly with 10s of billions (American) of air miles flown with this accident representing around 4000 miles of that or a one in 7.5 million accident rate based on air miles. Even if nothing substantive is done will it materially change the aircraft's safety record over its remaining life in the air?)


(Wikipedia says 789 planes. I figured 750 flying 18 hours a day 300 days a year at a piddly 400 MPH for 20 years - 32 billion miles flown. Heh, that's a bit more than the number of milliseconds in one year.)

Normal gee
 

@JD

We are talking about the aircraft reference frame for one gee. That's down from your brain thru your butt if you are not at Creech AFB flying a Predator drone 10,000 miles away, heh heh.

It's the same as when driving a car or flying a Cessna or riding on a roller coaster.

What makes this concept for FBW attractive is that the rate and acceleration sensors so essential to a FBW system do not depend upon earth-referenced gee or air data. Using body rates is also much closer to what we old fossils experienced in every plane we flew. You still have to use your attitude indicator and such when IMC, but somehow the old "seat of the pants" expression holds true.

The Airbus does not "trim" for one gee all the time, but corrects for attitude ( best I can glean from the manuals). So it trims for less than one gee when in a climb/dive. Also adds a bit of gee if in a bank. So in a level 30deg bank it trims for 1.15 gee +/- as the neutral position for the stick. The "other FBW jet" I flew didn't/doesn't do this. The gee command was always thru the body axis ( the jet and the human), plus we could trim the thing for something besides one gee.

Using a gee command system doesn't work real well in a video game or flying a Predator. You can't "feel" the gee.. So a body rate/attitude blend seems better for those applications. Have to check, but I think the Shuttle does that until the system switches over to air data at low altitude, prolly below 100,000 feet or so.

Questions?

If your plane is not "level" that could lead to a 1g through the butt accelerating auger in situation. (Although you'd be going gosh almighty fast pretty quickly.) Nose down and accelerating could end up feeling like one g through butt.

(And I'm only mildly surprised the Predator pilots don't have motion base chairs to simulate what the plane feels. One industry I do work for dotes on that sort of stuff as means of separating rubes from dollars. It's convention is in Orlando just before Thanksgiving.)

the report sugest it can also oscillate, and I hope that the amplitude will not reach to extrem values..... better rather degrease.... in case of the A 340 with that trust (set) it first climbed... the next question is, will it climb to the moon or at what level will it stop climbing? and will it be than inside the coffin values.....

but I think, first we have to draw a raw AoA line with: AoA = pitch-FPA and look how good the Mr.alpha-protector did his work

PJ2;

Thanks for replying. Yes, we agree completely on what you wrote in your #548 and #559, except for the phrase I quoted and which I may have misunderstood. The only thing I commented on, as a 'flight mechanics' guy, was the notion that "the reason for 'set power and pitch' is to 'disregard all those messages' ". And yes, the SOP and trained response has been along for a long time, but only after a series of mishaps, as I remember it. Somewhat similar to the insight that you have to reduce AoA to unstall an airplane.

Wasn't the THALES AA, which is/was the more prone to block at altitude, exclusively mounted on A330/340 ?

HN39;
Yes, agree. Hopefully the regulators, the sim builders and the airlines will continue the encouraging trend towards such awareness and changes in training. From a recent flight crew training manual revision, "High Speed Protection may also result in activation of the angle of attack protection. In all events, check the AP engagement status, and re-engage it when appropriate. It may have tripped and the associated aural warning may have been superseded by the overspeed aural warning.". I recall our original discussion on AoA in the second thread and how much I've learnt since then from discussions like the present one. The great difference between the stall AoA at low altitude and cruise altitudes was one of the few things that Davies did not discuss, (I suspect because he would never have expected that line pilots would be anywhere near situations requiring that knowledge) and was never discussed as a flight dynamics matter in any training. Many thanks, HN39.

PJ - can you explain the logic of the angle of attack protection? I can understand why the system might want to LIMIT AoA to alpha-prot (desirable), but why design it to pitch up to it from level flight (bizarre)

From my armchair I have had fitted an old battery powered A/H on each instrument panel on my own ABs (they are cheap enough for me to have two). Whilst not as accurate as the full glass panel, working as it ought to. One or the other should be enough to to give me time to check the aircraft's pitch and bank
are not excessive, so I have time to check both my speed and altitude sufficiently adequately, using my GPS. And the check lists...

(We had to use the Astrocompass to check the compass's Deviation before we could land using the SBA. I cannot do that in my A or B, now.)

Some Tiger Moths had a spring loaded vane as an ASI. It was unheated.

http://www.pprune.org/%3Ca%20href=ht...0/683pn2v7.jpg
......

@grity

Congratulations at reverse engineering that FDR data, I guess there's no chance of a FOI request for the actual data after all this time - along the lines the 9/11 data some of which is know in public domain.

To cross-check your FPA calc I reckoned the peak climb rate of 6,000ft/min was achieved at CAS of 250kts = 45,000ft/min TAS @ 37,000ft. FPA = asin (6/45) = 7.6 degrees.

Is the difference just my eyes?

sensor_validation your 7.6 is a good middle value between my 0 and 15 in the climb

?!?

Pitch
 

thanks Grity,

I will modify my two extra A/Hs to flash a warning light if my pitch exceeds say 5 degrees but this should be modified to allow for lower altitudes - and I am trying to manage without Pitot or Static inputs to my armchair !

This paragraph is a key point in your last few posts.
And sadly, the way the UAS procedure is written is a true reflect on how neglected that procedure has been dealt by Airbus for a situation at cruising altitude.

As a matter of fact, pre AF447, how many times have we trained for UAS at cruising altitude ?
In my own experience ... never !

Following the Air Caraibes incidents, there was an apparent willingness to address the issue – That was in October 2008 – 7 months before AF447.

http://i45.servimg.com/u/f45/11/75/17/84/af447_10.png

2 years after AF447 – A suivre … did not yet materialize.

Your concept Do nothing is mainly correct but is certainly not suggested by the UAS procedure the way it is still written up to now.
The way you detail it, is, IMO, mainly your own interpretation, and not supported by the text. It is not clear what is part of the memory items, what is not, what leads to what ... ?

As a starting point, what does mean If the safe conduct of the flight is impacted ?
In my book, the safe conduct of a flight is always impacted, from start to end.

And if the procedure has to be divided in two different course of action, as a start, it should be clearly mentioned to relieve any ambiguity :

• If the safe conduct of the flight is impacted …
• If the safe conduct of the flight is not impacted …

Where is such unambiguous bifurcation point ?

BOAC;
Thank you for the question. I want to say something.

I am a mere bread-and-butter retired Airbus pilot who loves technical and human factors knowledge and who labours under a peculiar fascination concerning this aircraft which I flew and this design and its history. Years ago I heard ET's presentation on "Why We Designed Them the Way We Did" (and have the paper on same), and have alternately embraced and puzzled over the airplane since first flying the A320 in 1992 and the A330/A340 in 1999. I have been educated by those with far more experience than I on the Airbus and by some engineers who are possessed by an unusual patience and ability to explain things to a pilot.

Though I would like it to have been otherwise, I know nothing of aerodynamics or computer control systems. I know something about flying the Airbus and its "habits", and how to stay alive flying airplanes, a bit about flight safety and organizational factors in accidents and a bit more about flight data analysis.

The singular thing this thread has taught me is how little I really know, and what a pleasure it is to learn from others, a relationship which I hope is reciprocal. The older and greyer I get the more willing I am to let stuff lie in the weeds because this stuff always surfaces. I am less willing to "pronounce" on the why, the how and the what in sophisticated systems into which I have the merest cracks to peer through. On staying out of trouble in the Airbus, I think I can say something of value and have. But, despite the usual conspiratorial crack-pot views, for easily understood reasons I lean to trusting the BEA on this one and am willing to wait to see what they have to say on various points so intelligently raised but frustratingly pondered here.

My views on the AoA item?...I think its "ordinary" vice "extraordinary". I would guess that in this AoA Protection Law we are simply looking at a response to the standard right-hand side of standard buffet curves, I can only posit that the airplane would pitch up "on the Alpha Prot AoA" in accordance with the design, perhaps with input from the High Speed Protection Law, who knows? We know from the AAIB report (if not from access to AMMs, AWMs & TSMs, etc), that when the Alpha Protection Law is invoked due to high Mach No., a pull on the stick would take the aircraft (in Normal Law) to Alpha-Max and no further. The risk of an accelerated stall is assumed to be low to nil in Normal Law. We also know that the pilot can intervene fully, with a stick-forward (ND) response to essentially cancel the AoA Law once below the Alpha Protect Law triggering point. Rather than second-guessing the designers and the engineers on each point and at every turn, perhaps thinking about why the design is the way it is in toto to see what problems are being solved? I think it is fair to say that some credit after all, must be given these people who, with the exception of one or two who have contributed to the thread, know far more about the design. Would we have the aircraft pitch down as per the low-speed Alpha response?...I don't think so. I think we have to think about it a bit, in the context of the Airbus design and not in the context of non-fbw designs and expectations.

To me, the AoA response makes complete sense as a pilot of the Airbus. What has not been either appreciated or grasped by many, is, from the beginning, this design has required a different way of thinking, a different approach to training, to flying, to anticipating behaviour and to interacting with the machine, and it absolutely requires a positive attitude and approach towards it. If one is forever either fearful or hostile or otherwise actively diffident towards the design and the airplane, like prejudice itself, no amount of discussion, facts, knowledge or perhaps even actual engagement and understanding, is going to change one's mind. And that too, is a human factor.

Would it be fair to state that the AB pilot, except for direct law, is really flying the autopilot and not the airplane? Assuming, of course, that the AB FBW system, with its various protections, is really an autopilot with several different interfaces.

It seems to me that, once they stalled, the proximate cause of their not realizing they were in a stall was the fact that the stall warning behaved badly. Put simply, if the aircraft knows that AoA is above the stall number, the stall warning should go on and stay on until the aircraft is not stalled any more.

What I know about FBW I have learned from PPRuNe. Whenever there is a discussion about a problem with flight control systems, it always seems to be about A and not B.:confused:

Understanding Angle of Attack.
 

PJ

I too am an old retired Boeing and Airbus pilot. My take is that AoA is important when dealing with very slow flight, i.e. when a Navy pilot has to bring it in on a very short deck with rolling seas. Or an F-16 pilot is carrying assymetrical external stores which alters stall speeds and AoA and has to restrict “G” loads after bombing runs.
Air carriers uusally have more than sufficent runway lengths thus never have to be very slow. Thus “Stall speeds and approach speeds” are pertinent. Therefore merely keeping a sufficient extra speed eliminates any need to even consider AoA.
And we have never had to even consider or train for it. In the plane or simulator.

We are not taught how to use it and do not have an AoA indicator that I’ve ever seen.

In auto any mode it may be there but to me invisible.

Here in 447 a plane runs into wild gyrations. Everything kicks off. The only task is to maintain control of the aircraft. Nothing seemed to work, not any of the automated protection systems. And zero guidance for the crew.

It is not possible to have any human train for such an eventuality. Nor for any human to cogently understand and deal with this myriad of failures all at the same time.

Other than to avoid getting close to this perilous situation.

All of the wise council regarding, “Well look at how many times pilots did not encounter this situation”, is not good enough for me.

Going into or near violent thunderstorms and their unknown qualities carries a risk that a paying passenger does not want, nor need to put up with.

PJ - you will find on another thread that I endorse your views that a " different way of thinking, a different approach to training" is needed and that I do not think it is in place.

I firmly believe that in this transitional period where humanity is coming to terms with a technology that is advancing at an incredible pace, aviation design should in simple terms try to conform to 'the norm' ie what would the average competent pilot do here? I have an overspeed situation - would I pull up 4000ft? No, I would control my speed another way. I am stalled - do I expect my warning system to quit? NO. I would like it to carry on telling me I am stalled - it is no secret - it knows! The system has trimmed my tail to the point where I may not be able to fly out of it. Just as with the AMS 737, I would expect some 'big numbers' sign to warn me and preferably to be asked if I really wanted that. Not '"Don't ever touch that wheel - you don't need to". That is what I mean by the 'logic' of the system.

Many years ago I began watching Airbus develop the fbw technology in a state of geeky technophile awe. The shine has long since departed. Friends who have 'crossed over' tell me it is a great system. A great system when all is going well is the kindest thing I can say.

Hi,

So .. we return to starting blocks ... again
If you know that (a different approach to training) ... why (it's seem's by all I read here from professionals) .. that the training is qualified as "low" at AF and certain other companies ?
If you know that a different approach is needed .. certainly Airbus and consequently .. companies know that.
Why .. after sooo many years .. some voices for more training and revised procedures .. ?
Who was blind or negligent ?

What would the average competent pilot do here?
 

A competent pilot driving in darkness toward known convective activity would have his eyes glued to the attitude indicator and occasionally cross check other indicators. As soon as the autopilot clicks off a competent pilot would take a deep breath and look about for possible reasons, but not do anything unless the flight path of the aircraft is changing. If the aircraft had been in trim before autopilot disconnect, there would not be any major departures from the previous attitude, unless there is a flight control system hiccup.

I have read, and reread, the BEA note over and over and I do not see any direct statement that the pitch up of AF447 was uncommanded. If someone else sees otherwise, let me know.

Someone earlier stated that, ordinarily, an autopilot disconnect should not be considered an emergency yet it appears that many AB pilots think it is. Clearly the AF447 pilot thought that way and he acted inappropriately.

BOAC, "The system has trimmed my tail to the point where I may not be able to fly out of it."

The system includes the pilot. And it was direct pilot input that caused that trim setting to exist. I also note it MAY be time for a remedial read on the BEA release again. The trim stayed sane until after the PF input full NU (and full left.)

Do a search on B737 RSEP.

Smilin Ed
" If the aircraft had been in trim before autopilot disconnect, there would not be any major departures from the previous attitude, unless there is a flight control system hiccup.

I have read, and reread, the BEA note over and over and I do not see any direct statement that the pitch up of AF447 was uncommanded. If someone else sees otherwise, let me know."

Then what caused the 7000 fpm rate of climb? If not strong updrafts. Which can cause attitude changes, and thence auto trim inputs.

The nose may be in a nose down attitude at the same time it is carried aloft.

Hi,

DCA09IA064

BOAC;
Well, I guess there is no point in continuing then. You have your mind made up re Boeings, in the face of a very successful airliner series which works very differently than the old B737.

The world is apace with technological change - has been for fourty years. Although you say philosophically that you 'firmly believe in this transitional period...', you shrink the argument to Airbus system design, resisting every point of explanation regarding the "new" technology offered by those who actually know and fly the machine.

The AF 447 accident is not going to be solved in this fashion, and the A vs B is never going to be resolved, but merely surpassed.

Thanks for the opportunity in our dialogues to do a lot of learning about both my airplanes. I loved the Boeings, Douglases and Lockheed in which I spent thousands of hours and when it came to move to the left seat on the A320 I can tell you it was very hard work and not many people knew very much about the airplane, (we didn't even have VNAV). I wouldn't take the Airbus "over" any other aircraft, (my favourite by far was the L1011-500) but it and the other two were a joy to fly and puzzle over, for fifteen years. For me the shine never left but I can understand why it might.

However, I think our concern should be for a new generation coming into the industry who "never flew an airplane" and were, as one A330 captain recently wrote about how to think of the A330, 'gamers' (not pilots), and I suspect there, we would find a few evenings over ales!

http://s7.directupload.net/images/110701/gw8vgrx3.jpg

the newer version with better CAS/TAS correction, thank you sensor_validation
of the FPA and AoA for the A340....during the zoom climb

with the given values of Altitude and CAS/TAS for every 5 sec. I calculated FPA (arcsin FPA = delta altitude / distance in 5s) yellow line;
and then with the given Pitch value the AoA (AoA=Pitch-FPA) blue line

the red and black line for Altitude and CAS has no scale....

if the AoA-protection-law was in regiment in the time between the MMO-exeed and 2 s after the first push with the sidestick, then the graphic show that in this time the AoA was not higher than 5 deg

"the AoA protection law seeks to hold the angle of attack constant at alpha prot" wrote the AAIB Bulletin.....

AoA is not the angel for the wing-flow, it is the angel between the direction of the fuselage and the flightpath (FPA) and the calculation did not include the turbulences.....

(I take away the first version)

This is precisely what the manual advertises ... but live experiment prove it wrong.

http://i45.servimg.com/u/f45/11/75/17/84/af447_11.png