AF 447 Thread No. 6

[john_tullamarine]

AF 447 Thread No. 6

Thread part -

(a) #1 starts here (http://www.pprune.org/tech-log/395105-af-447-search-resume.html#post5303737) and finishes here (http://www.pprune.org/tech-log/395105-af-447-search-resume-195.html#post6408432). Posts = 3890
(b) #2 starts here (http://www.pprune.org/tech-log/449639-af-447-search-resume-part2.html#post6408428) and finishes here (http://www.pprune.org/tech-log/449639-af-447-search-resume-part2-127.html#post6476460). Posts = 2537
(c) #3 starts here (http://www.pprune.org/tech-log/452836-af447-thread-no-3-a.html#post6476336) and finishes here (http://www.pprune.org/tech-log/452836-af447-thread-no-3-a-104.html#post6515428). Posts = 2071
(d) #4 starts here (http://www.pprune.org/tech-log/454653-af-447-thread-no-4-a.html#post6515515) and finishes here (http://www.pprune.org/6561320-post1061.html). Posts = 1061
(e) #5 starts here (http://www.pprune.org/tech-log/456874-af-447-thread-no-5-a.html#post6561270) and finishes here (http://www.pprune.org/tech-log/456874-af-447-thread-no-5-a-99.html#post6638007). Posts = 1978


Links to the various BEA reports are given below. If I have missed any of the useful papers, please PM me with the URL and I can include it.

(a) BEA site - French (http://www.bea.aero/fr/index.php), English (http://www.bea.aero/en/index.php)
- Report link page - French (http://www.bea.aero/fr/enquetes/vol.af.447/vol.af.447.php), English (http://www.bea.aero/en/enquetes/flight.af.447/flight.af.447.php)

(b) Interim Report (No, 1) Jul 2, 2009 - English (http://www.bea.aero/docspa/2009/f-cp090601e1.en/pdf/f-cp090601e1.en.pdf)

(b) Interim Report No. 2 Dec 17, 2009 - English (http://www.bea.aero/docspa/2009/f-cp090601e2.en/pdf/f-cp090601e2.en.pdf)
- Update Dec 17, 2009 - French (http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.17.12.2009.pdf), English (http://www.bea.aero/fr/enquetes/vol.af.447/update.on.the.investigation.af447.17.12.2009.en.pdf)

(c) Estimating the wreckage location Jun 30, 2010 (http://www.bea.aero/en/enquetes/flight.af.447/phase3.search.zone.determination.working.group.report.pdf)

(d) Wreckage search analysis Jan 20, 2011 (http://www.bea.aero/fr/enquetes/vol.af.447/metron.search.analysis.pdf)

(e) Briefing and associated update May 27, 2011
- Briefing (http://www.bea.aero/fr/enquetes/vol.af.447/info27mai2011.fr.php) - update French (http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.fr.pdf)
- Briefing (http://www.bea.aero/fr/enquetes/vol.af.447/info27mai2011.en.php) - update English (http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.en.pdf)
- Briefing (http://www.bea.aero/fr/enquetes/vol.af.447/info27mai2011.de.php) - update German (http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.de.pdf)
- Briefing (http://www.bea.aero/fr/enquetes/vol.af.447/info27mai2011.br.php) - update Portugese (http://www.bea.aero/fr/enquetes/vol.af.447/point.enquete.af447.27mai2011.br.pdf)

(f) Interim Report No. 3 July 2011 - French (http://www.bea.aero/docspa/2009/f-cp090601e3/pdf/f-cp090601e3.pdf), English (http://www.bea.aero/docspa/2009/f-cp090601e3.en/pdf/f-cp090601e3.en.pdf)

Miscellaneous pertinent links -

(a) Airbus Operations Golden Rules (http://www.airbus.com/fileadmin/media_gallery/files/safety_library_items/AirbusSafetyLib_-FLT_OPS-SOP-SEQ03.pdf)
(b) ALPA FBW Primer (http://cf.alpa.org/internet/alp/2000/febfbw.htm)
(c) C* and Civil Transports - Cranfield (https://dspace.lib.cranfield.ac.uk/bitstream/1826/186/2/coareport9303.pdf)
(d) Longitudinal Flight Control Design - RAeS (http://www.raes.org.uk/pdfs/2989.pdf)
(e) Longitudinal Stability: Effect of High Altitude and CG - Boeing (http://boeing.com/commercial/aeromagazine/aero_02/textonly/fo01txt.html)
(f) pitot static system performance - USN (Pax River) FTM (http://www.aviation.org.uk/docs/flighttest.navair.navy.milunrestricted-FTM108/c2.pdf)

Search hint: You can search PPRuNe threads with a filter in Google by using the following search string example -

ths af447 site:http://www.pprune.org/tech-log/ (http://www.pprune.org/tech-log/)

This will search for mentions of THS in the AF447 threads of tech log only.

Just change the THS in the string to whatever you want to look for. This allows one to search for any term or phrase of interest throughout the threads.

Adding the site:URL end part is the magic that restricts Google to only searching in Tech Log.

This filter technique is absolutely wonderful and can be used generically to find things of interest in PPRuNe - appears to work OK in the PPRuNe search function as well.

[Graybeard]

Wow, the first to post on this thread...

I get the feeling the Airbus 320+ was designed for single pilot operation from the beginning.

[Zorin_75]

Just a remark on the last post of thread #5 -
Predator Drones are of course flown by pilots, here's the "flight deck":
Image Viewer images/stories/full-size/uas_15-110712-02.jpg – TechNewsDaily (http://www.technewsdaily.com/templates/technewsdaily/image-viewer.php?img=images/stories/full-size/uas_15-110712-02.jpg)
It just happens to be not on the aicraft.

[Lyman]

takata re: SPIKES.

The Airbus has this elegant escape from CFIT? NON?

Max power, roll full, and pull back max.

A max effort, 'at the limits' safety manuever. During which the a/c nibbles at Stall? and wing drop? No sweat, the Bus knows SPIKES.

The Pilot doesn't. As above, for Alternate Law, the STALL WARNING needs some looking into.

[HarryMann]

Seems a naive viewpoint to me... It's not naive at all. Plenty of aircraft have very unpleasant stall/departure/spin characteristics, which no-one in their right mind would want to explore outside a very carefully managed flight test environment. That isn't to say they can't be operated safely, just that their pilots need enough warning to be able to stay away from the stall in the first place.
Fair enough HeavyMetallist, I didn't make a very good job of saying that it just isn't always possible to avoid the stall... we've had two or more in A-Buses fairly recently. And note too, this aircraft configuration seems to have fairly benign stall characteristics, stays pretty straight - waggles its wings a bit - but doesn't appear to spin or nod seriously. Likewise, from the Turkish Amsterdam experience, that 737 seemed to mush straight in, and was close to a recovery should some more height have been available.
Note! Both to some extent auto-trimming + inattention accidents.

But please don't think you'll ever avoid stalls 100% - that theory has been proved wrong since the Wrights. In fact, may well have helped cause this accident..
So.. at least... let us talk seriously about educating pilots about what stalls [really] are, not to be so scared stiff of them that they yank the stick back and open up the gas every time the very idea enters their head rather than a firm & steady ND, which has always been the No. 1 lifesaver! (which rarely will do any harm, speed is safety, even near the ground it can be fairly readily re-converted to PE!)



I personally believe the instumentation is poorly conceived... position error should be minimised 'by physical design' not using PE corrections (e.g. probes far fwd away from pressure field around wing or fuse, as in test fl;ight a/c). Actually the engineers designing these aircraft aren't stupid, and go out of their way to position static sources where the inherent pressure error is at a minimum over the normal flight envelope of the aircraft - those probes and static plates aren't where they are for convenience. Sure you can do better with a massive probe on the nose or trailing a static cone from the fin (or presumably several for redundancy http://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/eek.gif), but why got to all that trouble and expense when you can get acceptable accuracy by applying corrections for AoA etc?

I know they're not stupid ... I worked alongside some of them for years. But yes, again, was over the top in trying to make the point that the instrumentation aspects of this accident are not simply (and only) about icing pitots. Many assumptions that were made at original design time have to be re-assessed.. The whole thinking behind some of the most important basic instruments should be looked at again... it can very often be the 'off-design' case that suddenly become 'on-design' to catch you out. Here we have ASI basic (heating icing venting) pitot problem, ASI high alpha position error uncertainty, compunded by an AoA vane that really should be 99.99% foolproof and not reliant on an ASI cutoff, nor prone to any weather related problems itself and from Perpiganan - a lot more robust in basic nature.

Better backups should, indeed must, be de rigeour and not 'optional extras', ne'st pas.. these are civil airliners carrying many hundreds of passengers... lets just thank our lucky stars that many more have not dies in the last few decades from glaring faults that have caused many tens of extremely serious in-flight incidents.. every one a serious accident or a/c loss in the making.

and again, come back from the precipice of l/h and r/h sidesticks. They are NOT a good ergonomic solution

[Clandestino]

But please don't think you'll ever avoid stalls 100%It's type specific. First time I stall my current aeroplane, remaining lifespan of my passengers, my crew and me will be measured in seconds. 200 is very best we can hope for or perhaps it would be: the shorter, the better. I can understand the irresistible temptation to indulge in aeronautical generalizations based on single or few cases as I'll be first to admit it's boring, very uncool and perhaps even a bit frightening to rationally debate positively lethal stall characteristics of some T-tailed turboprop regionals and ways of dealing with them. Here we have real-life-case-study of heavy jet that even didn't spin at extreme AoA so we can indulge in heavily romantic notions such as: "they could have recovered it if: computers let them, they had yokes, stall warning didn't stop, they didn't fly into the storm (they didn't anyway), protections were working, it wasn't Airbus" or whatever half-informed mind would come up with.

and again, come back from the precipice of l/h and r/h sidesticks. They are NOT a good ergonomic solution Personally I find them acceptable. I have flown yokes with right and left hand, centrally mounted stick with right and left hand, sidestick with right hand, didn't find anything difficult or unacceptable about any of them and loved to handfly any aeroplane I was given at the time. A320 was by far nicest and easiest transport aeroplane I've flown though to be fair, I must add that it is also the only jet I'm rated on. She was docile even in direct law although I haven't experienced it outside the sim and I've come across only one pilot who got it on line (dual radalt fault). He claims it was even easier than sim and knowing him, I'll take his word for it.

It would seem that Airbus, Lockheed Martin (ex General Dynamics), Dassault, Sukhoi OKB and Cirrus are not believers in ergonomical unacceptability of sidestick.

[PuraVidaTransport]

I've made my views clear I think the primary cause of this accident is the Air France training/hiring/promotion practices however, I find myself thinking of some other things as well. The engineering of the aircraft, placement of probes and such is without doubt very good (and I am a Boeing fan). In almost all cases, things will work as they should and even (most) did in this case.
I do see valid points with the sidestick vs. yoke argument. I see a lot of times in the CVR/FDR data where the yokes may have made a difference. I think the yoke would have even more explicitly told the PNF just how badly the PF was 'over-correcting' for small deviations and he might have put some 'gentle pressure' on the yolk to smooth things out a bit. PNF had the experience, the seniority and the hours in the airplane but not the authority it seemed. With a yolk, 'sending a message' might have been easier through feel than verbally as he tried to several times unsuccessfully.
If this accident happened even in that case, we may have seen traces more similar to Egypt Air depending on how far the pilots were willing to push what they thought was happening. Looking at the last few seconds of the traces, strangely similar to Egypt Air and I'm sure this was due to confusion and two pilots not trained in how to react and their responsibilities in a crisis and a captain's poor decision to leave the authority in less skilled hands.
I don't see the yoke vs. sidestick being an issue except in crew teamwork. With good training, it seems to have been proven either is a safe and effective way to handle an aircraft. However, in those crisis times when two guys have to work as a team, the sidestick may not be the best solution.
Then again, what would have happened in the Egypt Air case had it been an A330? The computer would have summed the inputs and they would have had level flight, once they got the engines restarted obviously?

[Mr Optimistic]

There's an awfully long discussion going on about this but (as an SLF) would a real pilot please explain how you could sit for minutes with a nose high attitude, good power from the engines, a descent at 10000+ ft/min and not at the very least suspect - or admit as an option - you were stalled. As it clearly did happen it seems to me that the complexity and general protections offered by the aircraft in normal operations are such that they (and the training) have distorted the mental perceptions and outlook to the extent that physical realities have been exstinguished.

[HarryMann]

It would seem that Airbus, Lockheed Martin (ex General Dynamics), Dassault, Sukhoi OKB and Cirrus are not believers in ergonomical unacceptability of sidestick.OK, lets not confuse the issue with military references, I'm not silly enough to suggest that a sidestick is not the solution at high lateral and vertical 'g' loads in a confined space. It's also rather ignoring the fundamental of left & right propensity, since there'd only be one in amil a/c and whichever it was, you'd get used to it...

Ever tried hoola hooping ?

Which is your 'normal' direction - clock or anti-clock ?

Are you as good at it clockwise as anticlockwise, can you keep it up both ways ?

Unlikely, and what you are comfortable with, definitely helps in an emergency - one of the few times you'll be manual flying at altitude soon.. an emergency.


The Cirrus stick looks like you could almost use either hand - an interesting half way house?


I am very sure that the s/s, whilst it may (should) be mentioned in the final BEA report as a contributing factor (in the confusion at least, if not in possible overcontrol and climb (UP) fixation)... will never be condemned nor changed in ABs

.. because very much on the whole, it is satisfactory for the purpose intended and the vast majority of AB pilots are used to it, proficient with it and would not let it affect their flight control, whether left or right seat.

All this is not the same as saying that its innate ergonomics and possible l/r positioning for PF (and/or PNF) did not come into the safe conduct of flight after 2:10:00 for AF447 - because very clearly, in this instance, it did!

Shutting eyes to any facet involved in such a serious aircraft accident is never a good idea... whether 2,500 other aircraft have that facet or not... or even 250,000 others.

[hetfield]

These poor guys put the sidestick

NU - no stall warning
ND - stall warning

For me it's like a nightmare.

"What would I have done in this situation ?"

The more I think about it, the more I feel angry about this poor design and totally sorry for all the lifes it took.

[rudderrudderrat]

Hi HarryMann,

Some interesting points raised there. It seems a shame that the AB cockpit is designed around mostly visual and auditory clues.

The pilot is deprived of sensory feed back from the other pilot's SS inputs, elevator displacement load due autotrim (and by Elevator Displacement by FCCs) and Thrust Lever position when autothrust is engaged.

When it the automatics dropped out during UAS, Manual Thrust was required with no "muscle memory" of where the TLs should have been before, and there is no elevator load feed back due FCCs. Roll is direct and much more rapid than it was in Normal Law.

Why can't the AP simply hold "ATT" like we used to have with CWS (control wheel steering) 40 years ago? It would have freed up a few more brain cells.

[JD-EE]

Military pilots may last a lot longer than transport pilots.

Isn't it the opposite happening already ... ?

Hardly - fully autonomous aircraft are not fit, yet, to handle combat landing for a C130 or some aspects of air superiority work.

Those are the roles I had specifically in mind with that post. Taking pictures and dropping bombs when in a complete air superiority situation is relatively easy. How many UAVs or AAVs get plinked from the ground?

[JD-EE]

Zorin 75, some fully autonomous drones are appearing. Their only level of non-autonomy are the high level instructions - go to (location), circle it at some radius at some altitude, take pictures, send back interesting looking pictures. The chain of instructions can include a series of way stops for pictures and other uses. Nobody's "flying" it with a wheel or joystick. It's more typewriter driven and handles all the details of avigation itself. This hit the news fairly recently - amidst some apocalyptic rhetoric and reference to the Terminator films.

[bubbers44]

I think hand flying using those brain cells and knowing how to do it would have prevented the whole fiasco. 3 degrees nose up, 85% N1 would be a good start.

[Owain Glyndwr]

RRT

Why can't the AP simply hold "ATT" like we used to have with CWS (control wheel steering) 40 years ago? It would have freed up a few more brain cells.
Forgive me, but isn't that exactly what the Alternate law does if left alone to fly the aeroplane?

So how does an airplane with a pitch-rate command or g command fly? Essentially, it
gives you attitude hold with controls free, similar to an autopilot's control wheel steering
feature. If you change pitch attitude and release control pressure at the desired attitude,
the system holds that new attitude because the FCS reacts to bring pitch rate to zero.
The airplane should fly nicely with pleasant control forces and precise attitude control.
Air Line Pilot, February 2000, page 18
By F/O Steve Stowe (Delta), Local Air Safety Chairman, Delta Council 16

[rudderrudderrat]

Hi Owain,

isn't that exactly what the Alternate law does if left alone to fly the aeroplane?
I'm afraid not.
In ALT LAW, Pitch remains stable provided there is no ss input, but roll is direct and will constantly need an input until the aircraft is trimmed correctly. It seemed to give the PF loads of work because he hadn't practiced the technique sufficiently.

The only place to practice would be the simulator (twice a year normally - Unless he had UAS of course)

[Owain Glyndwr]

OK - I had pitch in mind not roll - I fully accept the difference.

[Lyman]

So in AL2, the Pilot is presented with different 'feel' in two axes?

The a/c is a package, and given the PF has made some fundamental mistakes, should one additional challenge be DIRECT in Roll, and not in PITCH? The Rudder needed to be a 'suggestion' from Captain?

You know, the fundamental flaw here may have been a clumsy catch from Auto. That is not difficult to understand in Simulator? The cost per pilot, five minutes in a 45 minute syllabus? Twice a year? A Grand?

[Diagnostic]

Hi rudderrudderrat,
roll is direct and will constantly need an input until the aircraft is trimmed correctly.
Or unless there is turbulance - true?

It seems to me that the PF had to make some roll inputs immediately after the AP disconnected, due to the turbulance (page 74, English version), and the (mainly NU) pitch inputs which led to the stall, were inadvertent and unintentional (e.g. perhaps muscular tension (not unexpected due to the "surprise factor"), and/or seat adjustment, and/or l/h vs. r/h seat usage, or other unknowns etc. as has been mentioned here before).

My point is that without having an "H-gate" arrangement (like a manual car gearstick) on a sidestick or control column - which I'm not advocating, of course - then it's impossible to only make inputs which are only in either roll or pitch. There will always be an amount (hopefully very small) of cross-coupling between inputs in those two dimensions (at least in my experience) especially when there is turbulance.

In other words, a design which requires roll inputs, is going to get some pitch inputs, like it or not, with a normal pilot - add in the surprise factor, turbulance, lack of high alt hand-flying training etc. etc. and this can all add to the amplitude of the pitch inputs, which seem to have been unrecognised by the PF, and the lack of recognition then caused the long duration of that NU input (which, integrated over time, drove the THS movement).

Can we ever expect a total lack of any pitch input, when there must be roll input, during conditions such as those? I'm uneasy (as a non-expert) with adding yet more automation into a situation like this, where the automation can "give up", but I can't believe there isn't a better way than dumping roll and pitch control onto a surprised pilot at FL350, at night, in some turbulance.

Interesting, the AB design seems to recognise a sort of "keep things as they were" philosophy (not unlike the UAS procedure, when above MSA) regarding the throttle at the point when the AP disconnected, since throttle lock occurs (until deliberate manual control of the throttles is commanded).

If that is appropriate for the throttle (and it seems to me that it is appropriate, for the short term immediately after an event like AP disconnect), why not also for the pitch control (again, until there is a deliberate decision from the pilot to takeover that function, which must not then be trained for pilots as an immediate reaction)? That would give some time for the PF to "catch up" with what is happening, rather than force him/her to takeover more than roll inputs initially... Just a thought.

[Edited to add: Of course that sort of "pitch lock" idea, would have to mean "with neutral elevators". Perhaps this isn't such a great idea, but I just don't believe that we can expect a total lack of unintentional pitch input, when there must be roll input, and with all the other factors that were against them...]

[Edited again to add: When you said earlier:

Why can't the AP simply hold "ATT" like we used to have with CWS (control wheel steering) 40 years ago?

Is that the same as having an initial "pitch lock" (until deliberately overridden, when the PF has "caught up") which I've been trying to describe (badly)?]

[3holelover]

My point is that without having an "H-gate" arrangement (like a manual car gearstick) on a sidestick or control column - which I'm not advocating, of course - then it's impossible to only make inputs which are only in either roll or pitch. There will always be an amount (hopefully very small) of cross-coupling between inputs in those two dimensions
I don't believe that to be the case. The stick's neutral position, in both axes, is fairly certain... it's rather easily maintained in that 'detent' - if you will - of one axis while being moved within the other.

[Machinbird]

In other words, a design which requires roll inputs, is going to get some pitch inputs, like it or not, with a normal pilot - add in the surprise factor, turbulance, lack of high alt hand-flying training etc. etc. and this can all add to the amplitude of the pitch inputs, which seem to have been unrecognised by the PF, and the lack of recognition then caused the long duration of that NU input (which, integrated over time, drove the THS movement).


A key element of manual instrument flying is building a mental picture of what the instruments are telling you and then controlling the aircraft with that information. With a PFD, it is much easier than with steam gauges since the majority of data is in a narrow visual span.

My present conclusion regarding the PF's control problems is that his scan was broken and disfunctional. His problems were very likely compounded by an inappropriate grip of the stick from the start, i.e. palm instead of finger tips.

The core of a manual instrument scan is control of the nose attitude and roll attitude-just put the pipper above the horizon line by the appropriate amount (~3 degrees in this case) and level the wings. Then drag in peripheral data from the sides, altitude, heading, and airspeed. Make small corrections as necessary. In the case of AF447, since airspeed was not available, apply suitable power to ensure stable speed.
In just a few minutes of this tedious flying, they would have been well down the road and away from the weather-but this did not happen.
Altitude never entered the scan, nose attitude did not enter (or else an inappropriate response to earlier training did). All the PF's attention appears to have been centered on controlling roll attitude which he was badly overcontrolling.

I can see two approaches to preventing this type of accident.
1. Provide a wing leveler function in Alt 2 law so the PF doesn't have to touch the stick except to maneuver. (The lowest common denominator approach)
2. Emphasize basic manual instrument skills under worst case conditions during recurrent training e.g. flying the S-1 and S-3 basic instrument patterns (by hand of course) at altitude in Alt 2 law. (You can always ask to try flying this just to prove to yourself that you can still do it.:})
I am hoping the regulators mandate the second approach.

[Diagnostic]

Hi 3holelover:
I don't believe that to be the case. The stick's neutral position, in both axes, is fairly certain... it's rather easily maintained in that 'detent' - if you will - of one axis while being moved within the other.
Thank you for that, and I'm happy to be corrected. I can only speak from my GA (control column) piloting experience, where no such neutral "detent" or bias exists. Now you mention it, the SS "detent" you mention, sounds rather like some PC joysticks :)

Having said that, I still believe that even with such a "detent", in a high-stress situation, the effects of adrenaline could easily overpower the PF's fine muscle control which is needed to "respect" such a bias toards neutral pitch. Therefore the only way to ensure that such a neutral bias is respected by a pilot who is "high" on adrenaline, may be to enforce it with some kind of lock. I'm sure that would bring its own set of challenges, however...

-----

Hi Machinbird, Thanks for your thoughts.

A key element of manual instrument flying is building a mental picture of what the instruments are telling you and then controlling the aircraft with that information. With a PFD, it is much easier than with steam gauges since the majority of data is in a narrow visual span.

My present conclusion regarding the PF's control problems is that his scan was broken and disfunctional. His problems were very likely compounded by an inappropriate grip of the stick from the start, i.e. palm instead of finger tips.
I've done limited IFR training, but your comments certainly make sense to me.

The core of a manual instrument scan is control of the nose attitude and roll attitude-just put the pipper above the horizon line by the appropriate amount (~3 degrees in this case) and level the wings. Then drag in peripheral data from the sides, altitude, heading, and airspeed. Make small corrections as necessary.
Perhaps, in the heat of the situation, with immediate roll inputs being required, this seems to bring us back to the PF being "behind" the aircraft and its initial state (inc. attitude), so that his NU inputs changed that state before he could get a good scan going (if that ever happened).

I also think the subsequent changing of what instrument data was available and valid (especially airspeed and v/s) could easily have led to confusion over which data to trust & which to ignore. Of course with hindsight, we can see there was "only" incorrect airspeed & intermittent loss of v/s indication - but he didn't know that :uhoh:

I can see two approaches to preventing this type of accident.
1. Provide a wing leveler function in Alt 2 law so the PF doesn't have to touch the stick except to maneuver. (The lowest common denominator approach)
2. Emphasize basic manual instrument skills under worst case conditions during recurrent training e.g. flying the S-1 and S-3 basic instrument patterns (by hand of course) at altitude in Alt 2 law. (You can always ask to try flying this just to prove to yourself that you can still do it.:})
I am hoping the regulators mandate the second approach.
Understood, and your approach (1) would have a similar result to my thoughts about a "pitch lock" i.e. the pilot would not have an opportunity to inadvertently change the pitch when controlling the roll, as he/she wouldn't need to control the roll either when the AP disconnected, with that "wing leveler".

My concern about approach (2) is that, while it is much preferable and I am all for pilot training for hand flying, it's different in the shock of an emergency. As someone else here said, a sim or training session where you know you're going to get emergencies thrown at you, or where you know you're not going to get them thrown at you, cause different human responses than a nighttime flight at FL350, with no warning that those emergencies are about to happen. Therefore can we really rely on approach (2), without also having some additional help (like approach (1))?

This type of situation where humans don't take over very well from a "monitoring" role, was discussed in a paper by Dr. Lisanne Bainbridge called "Ironies of Automation". Well worth a read, for those who haven't, IMHO. I'll try to find a link... This isn't specifically a piloting problem; it's a human problem and has also been seen at nuclear power stations etc.

[Machinbird]

Hi Diagnostic
One man's emergency is another man's "Oh the ding blab *@% autopilot just quit, guess I'll have to earn my living. Training is the difference. If those pilots had recent experience with the two mentioned instrument training patterns and could master them, then it would have been no big deal.

[safetypee]

The Ironies of Automation. (www.bainbrdg.demon.co.uk/Papers/Ironies.html)

The cognitive capabilities of humans. (www.humanfactors.uiuc.edu/Reports&PapersPDFs/chapters/Wickens_Durso%20Aviation.PDF)

Humans and Automation: Use, Misuse, Disuse, Abuse. (http://hfes.publisher.ingentaconnect.com/content/hfes/hf/2008/00000050/00000003/art00012) Free pdf download.

Trust In Automation: Designing For Appropriate Reliance - Research and Read Books, Journals, Articles at Questia Online Library (http://www.questia.com/googleScholar.qst;jsessionid=LHmZGvvpc1T1RT2pThDgGpvfwpMvh6f TTBPVC4hD8Tpg8J4LdQZy!-290722064?docId=5008765429)

The problem of automation: Inappropriate feedback and interaction, not over-automation. (www.jnd.org/dn.mss/Norman-overautomation.pdf)

Training for new technology. (www.crm-devel.org/resources/paper/bent.htm)

[Diagnostic]

Hi Machinbird,

Thanks again and I do agree that training is the key! I'm just not sure we can rely on some (beancounter-driven) airlines to give pilots all the training (inc opportunities for hand flying) which they really need - hence my wondering about what additional automation help can/should be incorporated, to avoid overcontrolling in Alt law hand flying. Perhaps the answer is "none"; but perhaps further automation help (in addition to more training!) might be useful.

Me, cynical about airline training policies? Surely not! ;)

-----

Hi safetypee,

Thanks for those links, including The Ironies of Automation paper which I mentioned.

[gums]

Thank you Safety, thank you. The Norman paper is appropriate and a necessary "read" for all here.

Before adding to 'bird's ideas, I must point out that the accident started when air data sensors went south.

So what do we do when that happens? What do we do when the A/P disconnects and autothrottle keeps power setting where it was?

Well, it depends upon the design of the flight control system (FCS), old fashioned or the new FBW systems. It also depends upon crew training and "attitude" of the crew.

As a pilot, I would prefer a more straightforward reversion sequence, similar to 'bird's #1.. No need to go straight to the "direct" law where electrons simply replace tubes, cables and hydraulic lines. Just revert to a very basic "control stick steering" (CSS) and replace insistent efforts by the FCS to "protect you". Provide warning and caution indications of mach, AoA, etc when available, but basically fly attitude that is available from embedded sensors in the FCS. In other words, you have a very capable autopilot type function and have time to assess the loss of air data while maintaining aircraft control. Retired, 'bird, Smilin' and others here have flown with CSS, and it is easy.

Inherent in a FBW system are embedded sensors we used to use for attitude and navigation/weapon delivery. When the confusers fail, these sensors also fail or are ignored, You are SOL at that point. Meanwhile, you should have body rate, acceleration and even attitude sensors embedded. e.g if. you don't depend upon the external navigation/autopilot sensors - inertial, GPS, doppler, pitot-static, AoA vanes/cones, AHRS, etc.

We don't need a "wing leveler", just a "hold the roll and pitch angle when control pressure/movement is relaxed". Throttle/power is manual. You can even have the gee command active and the roll rate command active when you move the stick, as they are inherent in the FBW system.

So reflecting upon Norman's outstanding discussion of feedback in automated systems, that's my story and I'm stickin' to it. He was too early for the AA regional ATR accident when the A/P was feeding in control deflection due to ice build up. When all went to hell in a handbasket, Sad sad, and there was also some distractions in the cockpit for ten or fifteen minutes before A/P disconnect.

Bottom line is there are too many "external" inputs to the FCS after the air data has failed, and the spurious stall warnings and such did not help the crew. With a straightforward reversion to a CSS type system, I think the crew would have handled the situatin better. I do not excuse the pilot's mysterious nose up commands for so long before the aircraft entered Chuck Yeager territory.

[Diagnostic]

Hi Gums,
Bottom line is there are too many "external" inputs to the FCS after the air data has failed, and the spurious stall warnings and such did not help the crew.
Agreed, sir - as others here have said before, a stall warning which turns off even when the aircraft is still stalled, is likely difficult for them to understand. I wonder how many AB pilots here, were taught that the stall warning could be silent with the a/c stalled, before this crash...? Yes, I know the attitude & airspeed should have been clues, but if they're (correctly) ignoring some instruments due to them being wrong (airspeed) or missing (v/s), how do they know they can trust the attitude displays?

Returning to your point about too many (pilot) inputs after the air data failed: Is it reasonable to expect civilian pilots who (whether we like it or not) now spend much of their time monitoring during the cruise rather than hand-flying, to instantly handle a totally unplanned transition to hand-flying, without occasional excessive inputs due to shock / surprise / fear / etc.? Especially with inadequate training for that situation? (Of course we're now seeing some improved training, as mentioned in the BEA report.)

My understanding of your eloquent posts (and please correct me if I'm wrong), is that your military flying was very different, in that you were very rarely using any form of (even basic) AP - yes? If so, then you were always "caught-up" with the exact state of your aircraft (through your ss!), in a way that a civilan pilot monitoring the cruise is much less able to be, unless hand-flying. Or do you believe I'm wrong?

It seems to me that the transition from "normal" to "problem" (it's not even necessarily an emergency, as Machinbird kindly pointed out), occurs to military pilots while flying; but occurs to civilain pilots, if in the cruise, when they are not (actively) flying and hence is a bigger shock and needs more time to catch-up, just because we're human. That's the time when control inputs have a greater risk of being inappropriate, due to the shock factor and lack of preparedness for that transition.

I think I'm probably just repeating PJ2 and his comments about them needing to "don't do something, sit there" :) - the requirement for immediate roll inputs prevented those initial moments of "catch-up", however.

I'm here to learn, and become a better (GA) pilot, so I'm open to being told that I'm full of &^%$ :)

[mm43]

An "in context" adaption of an excerpt from Ironies of Automation (http://www.bainbrdg.demon.co.uk/Papers/Ironies.html), by Lisanne Bainbridge, Department of Psychology, University College London.
Manual control skills
Several studies have shown the difference between inexperienced and experienced pilots making a flight path change. The experienced pilot makes the minimum number of actions, and the aircraft's flight path moves smoothly and quickly to the new position, while with an inexperienced pilot it oscillates round the target value. Unfortunately, physical skills deteriorate when they are not used, particularly the refinements of gain and timing. This means that a formerly experienced pilot who has been monitoring an automated aircraft's flight parameters may now be an inexperienced one. If he/she takes over he/she may set the aircraft's trajectory into oscillation. He/she may have to wait for feedback, rather than controlling by open-loop, and it will be difficult for him/her to interpret whether the feedback shows that there is something wrong with the system or more simply that he/she has misjudged his/her control action. He/she will need to make actions to counteract this ineffective control, which will add to his/her work load. When manual takeover is needed there is likely to be something wrong with the automated process, so that unusual actions will be needed to control it, and one can argue that the pilot needs to be more rather than less skilled, and less rather than more (sensory) loaded, than average.The above is just one of many factors mentioned in the above paper that will have had a bearing on the outcome of AF447. The inability of the PF to get immediate stable control of the aircraft was reflected in his actions throughout the final stages of the flight.

Once stalled, the phugoid nature of the aircraft's motion coupled with a similar rolling pattern appears to have resulted in a sensory blockage as to where the aircraft was in the flight envelope, let alone where it may have been - a direct result of factors mentioned above.

Physiological and psychological aspects of this accident will feature heavily in the final report.

As an example the "circadian rthymn" has previously been raised in the MAK/IAC Final Report (http://www.skybrary.aero/bookshelf/books/551.pdf) into the crash of an Armavia A320 EK-32009 near Sochi Airport on 3 May 2006. An adaption of a graphic presented in that report is reproduced below and attempts to quantify the increase in error rates expected at the low point of the circadian cycle which is nominally 0300 LST.
http://oi52.tinypic.com/2dbtpgw.jpg
The effects of time zone changes in relation to circadian time is difficult to determine, and in the above graphic the circadian time is shown as Local Solar Time. The social aspects of time zones that are quite removed from normal circadian time, e.g. Paris Local Solar Time and Central European Summer Time may have a modifying effect. Hence the showing of the Paris LST.

P.S. Thanks for the link safetypee.

[jcjeant]

Hi,

Will be interesting to have a circadian rhythm study about this accident:
http://www.airdisaster.com/reports/ntsb/AAR75-13.pdf
See page 17 .....
I doubt the circadian rhythm study will be useful ...

[glad rag]

What on earth are you on about.

[HarryMann]

Because the use of attitude references is a fundamental of instrument flying which is stressed in Northwest's flightcrew training program, the Safety Board concludes that the flightcrew improperly relied on airspeed indications as a means of determining aircraft performance.

Back to the importance of a scan including attitude rather than airspeed when in any doubt of a/s.
A clear AoA display is another thing - some have suggested here that it is probably yet to be accepted as primary display data

[jcjeant]

Hi,

What on earth are you on about. Check page 17
http://www.airdisaster.com/reports/ntsb/AAR75-13.pdf
the pilots actions and behavior and then .. think about the pilots actions and behavior of the AF447
Do you not see some similarities other than "circadian rhythm" ???

[Machinbird]

A clear AoA display is another thing - some have suggested here that it is probably yet to be accepted as primary display data

That is easily fixed in a few simulator sessions. Take away the airspeeds and then fly a few approaches.

Miracle of miracles, it works!:ok:

Properly set up to adjust for gross weight and configuration in the display, you won't even have to remember any numbers.

[gums]

Thanks for nice words, Diagnostic. All of us can improve our piloting skills, regardless of the thousands of hours and hairy moments.

My understanding of your eloquent posts (and please correct me if I'm wrong), is that your military flying was very different, in that you were very rarely using any form of (even basic) AP - yes? If so, then you were always "caught-up" with the exact state of your aircraft (through your ss!), in a way that a civilan pilot monitoring the cruise is much less able to be, unless hand-flying. Or do you believe I'm wrong?

I flew mostly single-seat planes. Had about 400 hours in a true two-seat interceptor. That plane had attitude hold, heading hold/select, mach hold, coupled A/P to the steering for attacks/missile launches ( and I used it once when firing an actual A2A roket - the Genie). Also could couple the A/P to the ILS for instrument approaches and "monitor", as many airline folks seem to do. My later jets had decent A/P functions and being the only soul onboard, I used it a lot when in IFR and had to switch approach plates or just get a grip on the situation. Of all, the Viper was the worse. Our FCS engineers resisted external inputs to the system. And to make matters worse, we had no warnings that the FCS was not accepting any further AoA commands for "altitude hold". So I lost a buddy over the glass surface of the Salt Lake as his jet gradually descended while he fooled with IFF codes and frequencies and such. The Airbus is the opposite philosophy of the Viper engineers, although I always thot we could have had a much better A/P.

You are correct, Diag, that when not relaxing and letting otto do the flying that we were much more aware of EXACTLY what our planes were doing. Nature of the business, and I would not expect the commercial pilots to fly at the edges of the envelope just to get from point A to point B. Just think about flying 13 or 14 hours over the ocean from the U.S. to Europe or Hawaii in your chair right now. No restroom or help from someone.. Did we use otto? You bet.

[Shadoko]

Question to pilots actually flying "jets": by day time, do one see the contrails of planes that are at the same level (so they would be seen "in the thickness direction")? If so, do it happens to fly "inside" them? If so, do it follow a particular smell of the air?

Questions to the "physicists"
- What can be the sectional area of contrails (I mean "primary" ones, not when they develop gradually when the conditions are right)?
- If the conditions are "favorable", could the flue gas of a plane which was "recently" in the same place (in the air mass, for failing to consider the wind) cause ice crystals (or conditions which could generate them) which may "ice" Pitots?

Well, questions may be stupid and irrelevant, but I was wondering if the increase in the occurrence of "Pitot events" could not be due to increased navigation accuracy, so that the planes on defined routes are all flying so exactly in the same place, that they may fly through the contrails of others if the wind is near front or back?

[jcjeant]

Hi,

Question to pilots actually flying "jets": by day time, do one see the contrails of planes that are at the same level (so they would be seen "in the thickness direction")? If so, do it happens to fly "inside" them? If so, do it follow a particular smell of the air?A trick used by german fighters attacking US bombers formations by the rear .. was to hide them self in the contrails of the B17 or B24 for approach at fire distance ..
I don't remember in any german pilot records I read the report of particular smell of the contrails or pitot icing (at a average altitude of 7000 meters) :8

[glad rag]

You wrote in your post

I doubt the circadian rhythm study will be useful ...

[RAT 5]

I've not read the whole thread, or report yet. However, there are many references here to lack of airspeed. Surely there was a ground speed readout. Thus there was not a total loss of speed indication. That info, coupled to basic power/attitude V/S could give some idea of what was going on. Easy to say from the arm chair, but I have tried it in the sim on TQ courses. OK, an approach is different to high level flying. Headwind/tail wind etc., but ground speed would change with attitude. Block the pitot tubes and fly an approach on groundspeed. It really emphasised that power/att V/S works. If they were confused by high nose no stall warning, low nose stall warning, then pause, consider and look for other clues. What does the ground speed do with change of attitude? That can at least confirm that the ASI is giving out B.S. At high level there is a lot of thinking time. Remember the BA 747 that had all 4 shutdown in volcanic ash over SE Asia. They had thinking time.
A scenario with loss of a primary parameter such as airspeed should be a recurrent training exercise. It has happened on enough occasions, resulting in a crash, to be included. In all cases the a/c was flying and controllable; either blocked pitot or staic vents. Sadly, many of the 3 year recuurency items are tick boxes. They are not related to real life incidents/accidents. If knowledge of these is not passed on and trained how will we learn and not prevent their re-occurence? After Hudson I know many airlines suddenly included loss of thrust and ditching. It was fun and ticked many boxes. The same after BA's B777 glider, but not Air Trans A330 glider. That would take too long in the sim. 30mins at least, but you would tick many boxes; flight controls, alt flap, alt gear, aircon & press, panic management, and it would be a great CRM exercise. Why was it not done before? Do we have to wait for another high profile crash to consider more meaningful recurrent training? With commercial pressure on crew resources training is diluted to minimums. The downward spiral is tightening.

[HarryMann]

Remember the BA 747 that had all 4 shutdown in volcanic ash over SE Asia.Indeed, but they had a Flight Engineer, remained disciplined, at post, and focused on trying and re-trying everything possible...
Flight Engineer worked tirelessly at re-starting engines, and eventually succeeded... whilst pilots flew, navigated, communicated and Captain thought things through strategically even, as well as tactically.
Even retained flexibility to change those plans at short notice, when engines started failing a second time...
Would in todays cockpit environs, IMHO, to be possible to lose focus and subsequently the aircraft, but then I am not flight crew, so have no right really to pontificate

[JD-EE]

To amplify just a tad on 3holelover's comment, there is a dead band around neutral position. That dead band makes all accidental inputs less likely. That's
the way I've programmed joystick based camera aiming software in the past. (The available joysticks were incredibly "noisy" around neutral position. So I had to put in a dead band.)

[JD-EE]

mm43, I seem to only be able to dream about a circadian rhythm that even approximately matches Earth's. I often joke I was born on another planet. Days about 28-32 hours long are ideal for me. And if I find myself working on a tough problem even that goes away. "Shop 'til you drop," is rhetoric, "Code til you drop," is reality, it seems. And when I do synchronize it's about "1200 to rise and 0400 to bed" in the local TZ no matter where I am. (SIGH!)

"That gurl ain't normal!"

Maybe they need to find pilots that can work well at night? (Probably there aren't near enough of them to matter.)

[syseng68k]

JD-EE, #41

"That gurl ain't normal!"

I dunno, sounds quite familiar to me :-).

Many years ago, before settling down to the requirements of family life,
kid's routines and defined working hours, I used to work around the
clock. Finished the day later, got up later, until I worked right round
to "normal" hours again. Something about solving difficult programming
or other design problems that really need long, uninterrupted sessions
to solve. Added to that, there's less noise and other distraction at
night, especially in the city.

Still have sessions like that occasionally even now, but not so
frequently...

[TangoLimaKilo]

Why not announce the measured angle of attack in the stall warning?

Then, after the warning, when in a stall, go into AOA (stall) ALARM mode.

Or, if AOA is past that of stall, why not have a voice that says, "Aerodynamic Stall"?

[Clandestino]

I must point out that the accident started when air data sensors went south. So what do we do when that happens?Use manual control to set and keep five degrees nose up, wings level attitude. Don't think about ergonomics, stick deflections, active control law or whatever else just use whatever control input is necessary to maintain 5° ANU. Set climb power. Ask your assisting pilot to dig out QRH and find attitude and power setting for weight, altitude and phase of the flight. Oh wait... that's exactly what "Unreliable airspeed" drill memory items tell you to do.

Trick is to have each and every airline pilot remember them and perform them flawlessly when his body clock says its 4AM, when he's jet-lagged, when he hasn't handflown the plane above FL100 for ages (if ever), when he does less than half a dozen manual landings per month, when he has never experienced control laws degradation and protection loss in real life and only seldom in simulator, when his knowledge of the principles of flight is lacking as he was only made to memorize the multiple choice answers to pass ATPL exams, etc... Whoever patents practical solution to this problem will quickly become millionaire.

He was too early for the AA regional ATR accident when the A/P was feeding in control deflection due to ice build up. When all went to hell in a handbasket, Sad sad, and there was also some distractions in the cockpit for ten or fifteen minutes before A/P disconnect.Simmons Airlines crash at Roselawn is a fantastic example of damage that zealous investigation can do to the efforts of increasing the flight safety. NTSB was after ATR and DGAC, BEA was set to defend them at all costs so went after pilots and in-between, many aerospace professionals who would benefit from understanding what actually happened at Roselawn were hopelessly confused.

A/P wasn't feeding any control deflection as ice was accumulating. Ice accumulation was light to moderate and did not affect the overall aerodynamics or performance of the aeroplane in any significant way. It was ridges of ice that formed on outboard parts of the wings, behind the de-icing boots that doomed the flight. It was result of flying with low speed and low AoA (feat made possible by extending the flaps in holding pattern to reduce pitch and improve passenger comfort) through area of supercooled large droplets - a largely unknown meteorological phenomenon at the time. These ridges destroyed ailerons' aerodynamic balance and caused hinge moment reversal at pretty low AoA - instead of centering under dynamic pressure, ailerons snapped to full right wing down position and that's what tripped the autopilot, surprising the crew and leaving them with roll control that needed very heavy force to move away from extreme position. Control was regained once, sadly lost again and when it was regained second time, there was no room to recover. Stabilizer and outboard portions of wings separated at 375 kts and 4G (IIRC Vmo on 72-200 was 230 kt). What got just cursory mention in NTSB report is that F/O tried to extend the flaps again as the aeroplane rolled, but as the speed was above limiting for flaps 15, flap extension was inhibited. Quite a reasonable feature that prevents blowing the flaps away now prevented recovery. Take note those of you seeking solution in more technology.

NTSB claimed that ATR and DGAC were aware of possible roll anomalies with ATR, yet they designed, certified and produced aeroplane as it was. Of course they knew about aileron hinge moment reversal, wind tunnel tests found it at 26° AoA, so far outside flight envelope that they never mentioned it to pilots. Bear in mind that ATR has very malign natural stall tendencies, it needs to be fitted with stickpusher to get certified and if you ever actually stall her, only luck can prevent your last flight from becoming hot topic on PPRuNe. During icing testing, no untoward effects of ice on controls were observed as there wasn't the method, and still there is none, for icing tankers to create SLDs so I really don't think there was conspiracy on ATRs and DGAC part to hide the dangerous side of ATR from her operators.

However, BEA played dirty. It tried to shift the focus from aeroplane to pilots and accused the crew of virtual dereliction of duty in critical phase of flight. That's where "distraction in cockpit" and "severe icing" myths mainly came from. Yes, there was cockpit banter but it was not out of ordinary and did not distract the crew from flying. The crew noticed ice accumulation and turned on de-icers. Their decision to hold with flaps was slightly unusual but not unreasonable as no one at the time knew that it could turn lethal. Our unlucky colleagues were test pilots without them ever realizing they were so. As Peter Garrison brilliantly summed up: Airplanes were picking up the ice since there were first airplanes. They get heavier and draggier, they propellers vibrate and engines overheat but pilots don't expected their controls to slam to stops without warning.

All you BEA bashers, please take no comfort or support to your fancy theories from my opinion on BEA's handling of Roselawn crash. IMHO, BEA is doing terrific job with AF447 investigation and there's quite easy to see whether report gets excessively biased: just check factual information against analysis, conclusions and recommendations to see if there's large discrepancy, which is good indication of foul play. One just needs a bit of aeronautical salt to do it.

[Lonewolf_50]

JD-ee:

How many UAV get plinked from the ground?

Well, given the dozens of various types and models, and operating parameters ... some are harder to deal with from the ground than others. I am guessing Global Hawk is pretty safe from ground fire. :\

For every innovation people eventually come up with a counter. :cool:

@ RAT 5
I've not read the whole thread, or report yet. However, there are many references here to lack of airspeed. Surely there was a ground speed readout. Thus there was not a total loss of speed indication.
This sub topic has been going on in each AF 447 thread.

The aircraft flies within the fluid medium, fluid being air. To see how important this is, try this exercise on your Ground Speed suggestion:

Fly a Cessna 172 at 100 kts IAS into a forty knot head wind.
GS ~ 60 knots.

Turn 180 degrees and try to fly at 60 knots using your ground speed as a reference.

Did you stall, or not?

It's much easier to do in a Jet Ranger, eh? :\

Even more interesting, take your Airbus A330 into an approach for landing at 155 knots, flying into a 40 knot head wind, using your ground speed for reference.

Do a touch and go.

Perform a tear type turn drop and then fly an approach to the same runway, opposite direction, using your ground speed as your speed reference ... and fly that 115 knots ground speed all the way to touchdown.

The landing might be firm, eh? :eek: Possibly short as well. :(

It seems to me that the point of the UAS drill in the A330 QRH is as follows:

There are charts for known best estimate pitch and power combinations, which will suffice (if you remember them) until you open the QRH and sweeten your pitch and power to match the closest number to "perfect" that will work to keep you in the safe operating zone until your pitot tubes, or you static ports, or whatever, unscrew themselves and airspeed indication returns to normal.

This "groundspeed" solution appears to me a solution in search of a problem, though I do understand the new feature in the A380 (an option) uses inertial inputs to aid in such situations.

From my PoV, over engineering.

[Lyman]

One such inequity. By Proxy, the designers of AB LAW were flying the LOC right along with the pilots.

The Pilot apparently misconstrued, mishandled, and became out of the control loop.

The LOOP was integrated into the airframe from the outset. It's shortcomings, no less then the Pilot's, are on display.

"Doing what was designed", is not a defense, it is an indictment.

[hetfield]

@Lyman

SPOT ON :ok:

[3holelover]

"Doing what was designed", is not a defense, it is an indictment.
No, it isn't. Like any machine, if it's mishandled it can bite.
If you built a bulldozer that someone drove onto the ice, and it fell through because the ice was too thin for a bulldozer, do you deserve an indictment?
If you built a car that someone drove off a cliff because he was going too fast for the corner, same question?

I won't keep going, because I know you get the point. The A330 is a terrific aircraft. ...and like every aircraft ever built, it can fall out of the sky if it's handled incorrectly. Sure, there are a few tweaks that could maybe help the next crew that finds themselves in the exact situation as AF447, and they're probably worth doing... But the biggest message here by far is:
- The lack of handling skills displayed by the PF,
- the evident lack of crew discipline or "professional" standards displayed by all crew members,
- and the obvious lack of training that allowed those first items above to exist, thus enabling the accident.

[hetfield]

- The lack of handling skills displayed by the PF,
- the evident lack of crew discipline or "professional" standards displayed by all crew members,
- and the obvious lack of training that allowed those first items above to exist, thus enabling the accident. To make a long story short:

Sidestick NU - no stall warning
Sidestick ND - STALL STALL

In addition, no tactile feedback at all....:ugh:

[Lyman]

Three

Let me describe a sitiuation, see what you think. The zig zag of the Pitch trace is straightforward.

The trace of ROLL is interesting. In DIRECT, 447's Roll is difficult to define.

Admittedly back and forth, and within reasonably close tolerances in both directions, the argument can be made that this ROLLING saved the a/c from a spin. Let's assume that is the result, since each excursion seems to have been met with a force sufficient not only to arrest the Roll, but to command a reversal. Perhaps divine intervention, no matter!

Who would lay claim to such a result? The Designers? Make her titchy so that Roll can be managed quickly? Not likely.

Or is it? She Stalled, and crashed. But she did not SPIN.

Theoretical, but discuss?

[DozyWannabe]

With all due respect Lyman, that's a poor excuse.

Several UAS incidents in the A330/340 preceded this one and none of them ended up in the drink, because in all cases the crew figured out what was going on and either applied the correct procedure or improvised one of their own to get them out of trouble.

If the design was so faulty as to take the crew out of the loop in every case then we'd be looking at more than one fatal incident due to LOC following UAS in an A330.

[3holelover]

What's to discuss Lyman? It's twitchy in roll when stalled? Please... That it mushed the way it did, rather than spin, may have as much to do with the somewhat aft CofG as it did the PF's valiant attempts to control it. Either way, the aircraft shouldn't have been stalled, and when it was, it should have been recognized.
For those two deadly faults, I feel rather confident pointing to things like AF's training, their SOP's, and their culture of standards and behaviour (which is likely encompassed within SOP's, but not ever having been a part of flight crew, I wouldn't know)

[ChristiaanJ]

Dozy and 3hole,
As JD-EE already said, "stop feeding the troll....."

[Lyman]

I am not now, nor have I said the design was faulty. It worked. It did not cause this loss. Please, please, slow down.

But to say it did not cause the crash is not enough. Neither is it fair to lump it all onto the Pilot.

It is not sufficient to design a system, sell it, and then neglect it. Beyond convenience, the politics are too cute. Teflon for all save the Pilot?

It was not supported well, and it was not supported in a myriad of ways.

Only a fool would repose in an imagined "perfection", to avoid responsibility of any kind.

Or a child.

For the record, I will stipulate that CONCORDE is an icon. All who have helped her live, and serve, should be honored. I mean it.

If one has something to say, one may say it.

3holelover. Having pointed at those two, do have some pointing left? What makes you think the Pilot, in Rolling, had any effect on preventing spin? I doubt it was so. He can't have been that intuitive......

DOZE You are correct in every way. There is less Truth in correct, than there is correct in Truth? Putting a stop to the discussion seems to be in vogue.

PJ2 He did NOT PITCH the a/c up to 15 degrees, at least not all at once.
Each Pitch Up was 'levelled off', as I read the Trace.

At Each Pitch up, PF may have felt a slowing, and NOT felt a continuation of NU.

IOW, He may well have thought that his Pitching Up solved an overspeed, albeit incrementally. Absent cues, he may have inferred an attitude closer to level than markedly NoseUp. The lack of THS Trimming on the way up shows us this. Several short and sequential commands for slowing. Instead, it was an ascent by degree, as I see it.

It is most unfair to propose that the Pilot merely wanted 15 degrees nose up. If that was so, the Ascent would have included the THS?

Throughout, the Pitch track is ragged, at least to me. Save until post STALL.

Stand to be corrected. And no shame in that, if by you.

[Clandestino]

I agree that just keeping attitude and power that were there when UAS is encountered is pretty good tactic for dealing with UAS. however, drill I was taught was quite good too and it's not add 5 degrees pitch, it's achieve and maintain 5° pitch above FL100 or MSA, which would make whopping 2.5° pitch up from usual cruise attitude, or at least that was the way I was taught on the bus. That attitude combined with climb power was calculated to keep one away from overspeed and stall so I find FCTM warning (we didn't use Airbus documentation, ours was airline tailored) about possible overspeed a bit worrying. PJ2, could you please post what are current memory items for 330 if crew somehow concludes that safe conduct of flight is indeed jeopardized by having unreliable speed?

It can be reasonably posited, though of course not confirmed, that the PF executed the pitch-up as a result of pitot failure and the memorized response, intended for the takeoff phase only. There is no other reason that makes sense as to why the PF would instantly pitch the aircraft up to approximately 15degrees.
Reminds me of misapplied tailplane stall procedure theory that was put forward when Colgan 3407 went down. Are they really plausible or are we're doing our best to distract our minds from the scary picture of startled pilots shedding the most elementary pilot skills and trying to climb away from something only they could see and force their aeroplanes where they were absolutely incapable of going?

[Lyman]

So then, to no one in particular. I will assume in the absence of response, that "He Instantly Pitched up to 15 degrees" is not correct. The a/c did not, and no one short of Nostradamus can say what the PF Wanted.

The stick was positioned aft. To assume the PF wanted other than to arrest a Nose Down (0degrees, less than level, in stable cruise) is conjecture, let alone a "command to 15 degrees".

Since everyone "Cannot for the life of me understand the PF'actions", one may ask how then they can be reported?

"The climb is not understrood". Fine. Absent volition, then, how is it one can proclaim an opinion?

There is a careering from one stop to the other here of logic, and the upshot is that PF gets painted with the fools brush.

At the very least, it is unknown, if not ill-considered, whether the aft stick was intended at all.

So we are pleased, then, that the thrust of the thread has moved onto "human factors".

Not in evidence, and dare I say, Pompous. Also dismissive.

The record is cherry picked, parsed, and incomplete, which makes it not only unreliable, but in opposition to any conclusion.

[bubbers44]

Nostradamus wouldn't have pitched up 15 degrees if they had airplanes then. Most pilots would have pitched up 2.5 degrees and used about 85%N1. This guy was out of control.

[Lyman]

I'll defer to your confidence, then. Bub, where are the Rudder tracks? I am not successful to enter the BEA website.

Cap in hands, can some kind soul post the Rudder traces?

Most thankful, Humbly

[GarageYears]

http://i1116.photobucket.com/albums/k579/GarageYears/AFF447_Controls.jpg

[oldchina]

I too had big trouble trying to get into the BEA site using Internet Explorer. Certainly no way to open their big report files.

In desperation I downloaded Google Chrome and can now get into the site. It's still a bit slow but it works.

[Rob21]

IMHO, at the present level of automation, I don't see enough reasons for an autopilot disconnection. If the solution for UAS is so simple (pitch & power), why the autopilot "refused" to hold the same pitch & power when they lost airspeed indications?

It amases me that a computer "logic" can not perform a partial autopilot disconnect, holding only attitude and power. Forget all the other stuff (navigation, altitude hold, etc...). An autopilot should not "panic" and disconnect just because of UAS. The computer still had enough data to keep the a/c flying, but it wasn't "programed" to do so...

Fill the cockpit with flashing and audio warnings, but give the pilots time to figure out what is the problem.

Autopilot is there to do exactly this: give the pilots time to do the thinking.

[ChristiaanJ]

IMHO, at the present level of automation, I don't see enough reasons for an autopilot disconnection. If the solution for UAS is so simple (pitch & power), why the autopilot "refused" to hold the same pitch & power when they lost airspeed indications?
Two answers from me, and I'll gladly stand corrected by the 'bus' experts.

- Most of the gains (and some of the logic) in the basic control loops (even in attitude hold) are functions of IAS (or, more usually, Mach).
For a designer it's not easy or obvious to pick a 'default' value to drop back to, when those air data go belly-up (NCD).... otherwise, if you think about it, he wouldn't have introduced those variable gains in the first place.

- The solution for UAS is not simply "maintain pitch and power", and neither are the optimum pitch & power the same values just after take-off or at FL350.
What's more, the actual pitch & power at the moment of the UAS first occurring are not necessarily the ones you want to maintain (depending on the moment in the flight where it occurs).

Finally, the autothrottle did leave the power at the last setting.
Whether it would have been preferable for the autopilot to have reverted to a basic attitude hold is IMHO very much an open question, especially in turbulence.

It amazes me that a computer "logic" can not perform a partial autopilot disconnect, holding only attitude and power..... The computer still had enough data to keep the a/c flying, but it wasn't "programed" to do so...I tried to explain above why it wasn't "programmed" for that.

Question to all : has anybody already worked out from the report, what the AP (long and lat) and AT modes were at the moment of the disconnect?

Rob's remark above about 'partial disconnect' reminded me of some systems I've seen, where partial input data loss would disengage the affected 'higher' mode, but without disengaging the AP itself, which would revert to ATT HLD and/or HDG HLD.

[gums]

CJ: I agree with Rob21, and I previously mentioned that the Viper had a "standby gains" function for just this reason - bug in the pitot tube(s), ice, physical damage ( and we had one accident from a huge pelican ripping away the radome and air data sensors).

Granted, we had a much greater speed range and less worry about mach "protection" , but the fixed dynamic pressure ratio value was a bit above normal cruise CAS/mach. It was much lower with gear down, as the stick would have seemed too 'stiff" for the amount of control surface movement we were demanding and the confusers would have applied less movement for the flight condition.

So this seems a very straighforward reversion function and should not completely eliminate the basic attitude control functions of the A/P. Lose the pitot system and the STBY GAINS lite comes on. Hmmmmmm, no big deal and the jet just seems "stiffer", more sluggish for "x" control stick input. The system could even use last known values, but most engineers would prefer a fixed value in non-volatile memory as we did. This implementation would also keep the "mach" protections outta play, but the crew would know this when the warming lite illuminated.

Further, there are probably pressure functions in the control actuators themselves that "limit" them as to rate and amount of movement depending upon actual forces exerted upon the control surfaces.

[CaptainGef]

Hello
If you read the graph p111 between 0211:00 and 0212:45
During all this time this is the stall warning the crew does not believe.
We can see Altitude does not vary too much and to quickly (37924 to 36068)
F/O pitch orders up and down.
Stabilizer from -3 to -13°
Speed is decreasing and as speed is decreasing, they don't see it and what does the THS do on FBW machine?: It tries to maintain flight path, so as speed is decreasing to maintain flight path the autotrim goes up to the limit, incidence is growing up to maintain flight path.
On a Boeing, even in alternate law, you would have to trim as speed decrease to maintain flight path, it would alert you that something is wrong.
Here, it is like if the pilot is being helped, without being conscious, to install the Airbus in the stall.
No sensation on the joystick, if he does not look sharply at his trim indicator, he does not see, he does not feel that the autotrim is trimming for him to install the stall !!
I think it is not the pitch orders that make the THS pitch up. It is the FBW system and this way does not help....

What do you think ?

[airtren]

For more thoughts on this, you can see Post #2934 on the Rumors AF447 thread:

"Post #2934 on Rumors AF 447 Wreckage Found thread" (http://www.pprune.org/6643908-post2934.html)

and Post# 1862 on this thread :

"Post #1862 on Techlog AF 447 Thrd #6" (http://www.pprune.org/6633045-post1862.html)

Hello
If you read the graph p111 between 0211:00 and 0212:45
During all this time this is the stall warning the crew does not believe.
We can see Altitude does not vary too much and to quickly (37924 to 36068)
F/O pitch orders up and down.
Stabilizer from -3 to -13°
Speed is decreasing and as speed is decreasing, they don't see it and what does the THS do on FBW machine?: It tries to maintain flight path, so as speed is decreasing to maintain flight path the autotrim goes up to the limit, incidence is growing up to maintain flight path.
On a Boeing, even in alternate law, you would have to trim as speed decrease to maintain flight path, it would alert you that something is wrong.
Here, it is like if the pilot is being helped, without being conscious, to install the Airbus in the stall.
No sensation on the joystick, if he does not look sharply at his trim indicator, he does not see, he does not feel that the autotrim is trimming for him to install the stall !!
I think it is not the pitch orders that make the THS pitch up. It is the FBW system and this way does not help....

What do you think ?

[Lyman]

So look, at high Mach, and AL, the RTL hobbles the Rudder.

I'll ask (rhetorically) again, if AB thought too much Rudder was a problem, why then did they not also think Pitch might need a looksee? TRIM?

Weary of this "dummy pilot" meme, I'll be brief.

If I am in Court, suing Airbus for this STALL accident, I need do one thing only. Hire a dolly, and roll in an RTLU.

"What is that", says the Jury. Case Closed.

:ugh:

[Rob21]

CJ's quote: "Finally, the autothrottle did leave the power at the last setting.
Whether it would have been preferable for the autopilot to have reverted to a basic attitude hold is IMHO very much an open question, especially in turbulence."

At the moment of autopilot disconnect, main concern is to fly the a/c.
In Alternate Law, some attitude help would be welcomed, IMO.
Reverting to the last attitude could have given more time for the pilots figure things out.

If autothrottle left the power at the last setting, why not have also an attitude hold at the last setting?

IMO, anything that can give more time for the pilots, is very welcome.
UAS is an emergency situation, and in a very complex FBW system pilots need time to sort things out. Autopilot should give a "hand", basic attitude hold. This will not bring an a/c down as fast as an inadvertent stall.

Just a thought...

[Lyman]

Hi Rob.

The BUSS system is available. Also available is an Artificial Horizon, independent, and gyro driven.

Air France chose not to fit these to this A330. To some folks, these additional, though optional systems, may also have been a help.

I think in retrospect, those who lost loved ones would have accepted a small hike in the ticket to have had this kit aboard.

Yet we see even this proposal dismissed, and in the name of what?

PRIDE.

#1 If it can happen, it will

#2 If it can be foreseen, and mitigated, it should be.

#3 The rest is autopleasure.

[3holelover]

Also available is an Artificial Horizon, independent, and gyro driven.

Air France chose not to fit these to this A330.

??????? Can you validate that comment? I've never seen ANY airliner without a standby horizon.....

I know I'm not supposed to feed any grumpy little three fingered bridge minders, but I'm having a hard time watching some of these comments go by. The RTLU hobbled the rudder? What on earth....:confused:?
I get the sense that you're prepared to point fingers at anything and everything except the truth.

[Lyman]

3holelover.

The Rudder Travel Limiter 'hobbles" the Rudder. It limits it from unsafe deflection into a too energetic airstream, hopefully preventing damage, or loss of Rudder and or Vertical Stabiliser. The Artificial Horizon is very definitely available as optional equipment, it is fitted on Captain's side, high/left, as I understand it.

BUSS (Back Up Speed System) is also available, and I think provided as basic on the 380.

ISIS is definitely Standby, but driven by the same system that powers the PFDs. If unavailable, a "Steam" AHI is nice.

Do you take seriously the accusation that I am Troll? I thought better of you. I am a pilot, been flying for 40 years, flew commercially, and at one time owned my own air freight concern.

I take flying seriously, I have never had an accident, and I frequently canceled if I thought there was even a chance of anything unseemly. I don't throw tantrums, or poop, and I respect everyone. Making silly accusations is not appropriate, imo.

If something disagrees with me, I may comment on it, I may ignore it, but to personalize it or stoop to childish harping is a waste of time, everyone's.

[Linktrained]

Lyman,

It would appear that AB sent a FOT saying that the BUSS system should not be used above F/L 25.0 on 9th. September 2009.
The A/H installations appeared in a photograph on an earlier thread, one for each pilot. ( I do not know how they were powered. I might prefer their own battery(s) !)

[3holelover]

Lyman, I know what the RTLU does. My "what on earth?" question is, how can you possibly be considering it in any way related to this accident?

I'm sure I've already noted, I'm not a 330/340 guy... I'm an AME, licenced on the 320 family, but the only bigger Bus I have on my licence is the 310... so I'm a little unfamiliar with the 330 panel..... I understand the beast had all the attitude info they needed, and that it was airspeed info only that they were missing for a short time. Am I wrong?

[Lyman]

3

I don't consider it related to this accident (RTLU). I didn't say it was.

I said that as a consequence of degraded Flight Law, the Rudder Limiter prevents extension beyond certain speed (calculated) limits, to prevent damage.

I compared it to a THS which in Alternate Law, is NOT LIMITED in deflection, and in 447's case, migrated to its NoseUp stop.

The comparison was meant to highlight the decision AB made on the one hand to protect a flight control from damage (Rudder/Vertical Stabiliser), and a disregard for potential Attitude problems caused by an unlimited TRIMMING DEVICE.

There are some well established pilots here who have noted this problem with autotrim.

Not only is this autotrim without designed limits, it works without notification to the cockpit. (Sorry, not correct, the pointer on the manual wheel shows the THS' deflection).

So this sets the stage for some (what I would call boneheaded) design that arguably is responsible in a very demonstrable way for the outcome of 447's flight path.

Did the THS prevent RECOVERY FROM STALL? It does not matter, strictly speaking. The fact that it deflected automatically into a position that caused potential threat suggests a serious flaw in the design of the Aircraft.

Now the argument can be had, (and is) that the PF was ___, ____,_____, and _______, but that is a dodge, intended to distract from the fundamental problem.

As to Attitude, the selfsame lack of design can be isolated and criticised. The Displays were "acting up", and one must admit to some possibility for the various displays to be unreliable, or transient, or gone. So an AH driven independently of the system that was in trouble seems a no brainer, at least to me. One would have to have a good argument against it to say it was not necessary. Clearly, if available, it MIGHT have helped.

[ChristiaanJ]

LT and 3hole,

- So far there is still no mention of the primary attitude data or displays going t!ts-up at any time, so discussing the standby horizon is just a sideline... (my firm made them, hence my interest).

- I know there were some photos, but that's the trouble with these forums with hundreds of posts each.... find them!
And not all A330 flight deck instrument fits are identical, so how to confirm you are really looking at the AF447 flight deck?

LT, in 'my' days, the (mechanical) standby horizon ran off the emergency DC bus via its own inverter, and it had its own 'high-speed' gyro, that would keep running on its own inertia for minutes even after electrical power totally disappeared.

I have no idea whether ISIS, or any other standby horizon fitted today, has its own little battery pack (lithium, four AAAs?) to keep the LCD display alive, 'when all else fails'.


Apologies for the O/T.

[Linktrained]

ChristianJ I was perhaps recalling the Hermes 4a, where the emergency standby instruments were from a constantly recharged battery ( 24 volt ? As a pilot I was not entrusted to use a voltmeter !) On checking, my Type Rating has now expired.

Precise instruments used to be backed-up by simpler, more reliable, if not quite as precise, earlier generation ones - think of any remote indicating Compass, which would be backed up by something that Captain Cooke would have recognised, subject as that would still have been, to more magnetic deviations and acceleration errors.

I carried a load of Rhodesian Riflemen to Singapore. They had not flown before and asked to see the Flight Deck of the Hermes. The plywood door was enough security, then. They reported back that we :
" had a hundred and twenty twelve clocks, all telling different times..."
which was poetically accurate - but I never counted to see if they were right.

[GarageYears]

Lyman said:

I said that as a consequence of degraded Flight Law, the Rudder Limiter prevents extension beyond certain speed (calculated) limits, to prevent damage.

So, probably of not much use then, given this accident was initiated by UAS (so no speeds) and then stalled (also little useful forward speed).....

Again, you refuse to accept the pilots did anything wrong? Right?

You incessant babble is tiring.

I want to ignore you, but you keep changing your username.... why so?

[3holelover]

Lyman,
The THS existing limits need no further limiting, and ought not to be limited any further, since there are imaginable scenarios which could warrant full travel. The rudder, on the other hand, had better be limited, lest air loads exceed structural limitations. Not comparable.

...and when the THS moves, it's not only it's little pointer on the scale that provides indication of it's travel.... there are big white stripes on the large black wheels inboard of each pilot, on the side of the pedestal. When the THS is moving, even peripheral vision will normally draw ones attention to those contrasting white stripes moving by.... Maybe the "whooler" shouldn't have been eliminated, but I'm not at all sure that would have made any difference here.

[Lyman]

I will defer to the certifications, 3holelover. My assumption is that Trim is not considered mandatory for certifying recovery.

That makes it optional. Throw in automatic, and one finds oneself STALLED with a trimset that "may be required in some situations."

Pass. If the a/c can get into a STALL, it should be able to get out the STALL with Elevators. ONLY. In this instance, the THS was NOT trimming the a/c into STALL. IT went all NU after the STALL. Do NOT now say, "BUT THE PILOT", because that is a distraction. It was STALLED, THS at -13, and that was that.

Should we continue to go around the maypole? "But the pilot put it there", YES. We shouldn't have to design a/c as if pilot's will STALL them?

Oh, yes, we should?

I do doubt that the THS is vital for recovery. If I am mistaken, I apologize.

If, and regardless of the pilot's skillset, the a/c is gonna STALL, pumping a TRIM WHEEL to get back to square one seems at the very least counterintuitive, if not demonstrably fatal.

[3holelover]

Should we continue to go around the maypole? "But the pilot put it there", YES. We shouldn't have to design a/c as if pilot's will STALL them?

Oh, yes, we should?

Forgive me Lyman, perhaps I'm thick, but I really do not understand what you're saying here??? Please, if you choose to answer me, make it clear for me, will you? Without your question marks further confusing me as to your point?

Also, re: "I do doubt that the THS is vital for recovery." :confused: ....so again, I don't understand why you're going on about it?

If, and regardless of the pilot's skillset, the a/c is gonna STALL, pumping a TRIM WHEEL to get back to square one seems at the very least counterintuitive, if not demonstrably fatal.Unnecessary. All he had to do was hold a nose down input and the trim would have followed. But counterintuitive? I should think trimming is no such thing to an experienced pilot! It's probably as close to naturally intuitive among pilots as lifting a wing when it falls. You said you were a pilot.... did you not retrim when you changed attitude?

[Lonewolf_50]

3hole, a few threads back, I think number 4, a gent who has taught in the A330 simulator related to us how some Unusual Attitude drills and recoveries required use of the trim wheel, manually, to get the pitch under control (in other than normal law, IIRC) in a suitable time frame. He wasn't sure if that was a "simism" or not.

Point?

There were or are some training drills that can be used to accustomize crews to using the trim wheel to control nose pitch in cases where the auto trim is either too slow or not helping redirect the elevators and THS where the pilot needs them. You could call the trim wheels a secondary flight control (Not sure if I am right about that) which means any rated pilot ought to know, like the flaps, when and how to use it.

And practice doing so.

[A33Zab]

STANDBY INSTRUMENTS D/O
ISIS
DESCRIPTION
The back of ISIS is equipped with two pressure connectors:
- one is connected to the standby pitot probe for total pressure
acquisition,
- the other one is connected to the combined standby static probes for
static pressure acquisition.
The ISIS indicator contains the following subassemblies:
- one pressure module,
- one inertial module,
- one computation module,
- one display module and,
- one interface module.
The pressure module is connected to the total and static pressure
connectors. Each pressure line is connected to a pressure sensor in
the pressure module. The inertial module is composed of three
gyrometers (gyrolaser) measuring angular speeds and two specific
linear accelerometers (pitch + roll). The computation module includes
a Central Processing Unit (CPU), which calculate the operational
parameters (attitude, altitude and airspeed), and a graphic card, which
receives the data from the CPU and sent it to the display module. The
display module is fitted on the front face of ISIS. It is of the Liquid
Crystal Display (LCD) type. The front face is also fitted with several
knobs for operation purposes. The interface module is composed of:
- a filtering board linked to an electrical connector at the back of ISIS,
- the 28V DC power supply unit,
- the interface board which links the pressure, inertial, computation
and display modules.
POWER SUPPLY
ISIS is supplied with 28V DC from the ESSential BUS bar. In case
of loss of this bus bar, the HOT BUS bar automatically takes over,
provided the Computed Air Speed (CAS) is greater than 50 kts. In
case of a power supply cut-off less than 50 ms, there is no effect on
the display.
OPERATION
When ISIS is energized, a delay of 90 seconds is necessary for
initialization. ISIS computes and displays own air and inertial
parameters. When the airspeed data is not valid, the CAS information
is provided as a back-up by Air Data/Inertial Reference Unit (ADIRU)
1 or ADIRU 3 regarding the ATT/HDG selector switch position. The
ADIRU also provides a Ground Speed (GS) information for
flight/ground condition, as a back-up. A BAROmetric selector knob
enables the display and adjustment of the standard barometric pressure
in hPa. When the Landing System P/BSW, located on the upper right
part of the indicator, is pushed, the G/S and LOC scales come into
view. The BUGS P/BSW allows to display the BUGS page. This page
is used to program characteristic speeds and altitudes displayed on
the related speed and altitude scales. The (-) P/B is used to get access
to the next bug and the (+) P/B to return to the previous bug. Air and
inertial parameters are processed by the computation module using
data from the pressure and inertial modules. The pressure module
sends the total and static pressure data while the inertial module sends
accelerations and angular rates data. Once computed, the parameters
are sent to the display module for indication to the crew. The LS data
is sent by the Multi-Mode Receiver (MMR) 1. Air Data/Inertial
Reference System (ADIRS) parameters computed by ISIS are sent to
the Flight Data Interface and Management Unit (FDIMU). The ISIS
can operate from 0 to 600 kts without deterioration of airspeed data.
It can operate from -2.500 to 55.000 ft and up to 40.000 ft/min without
deterioration of altimeter data. For the horizon data, the ISIS can
operate from -180 to +180 deg without deterioration.

http://i474.photobucket.com/albums/rr101/Zab999/ISIS.jpg

http://i474.photobucket.com/albums/rr101/Zab999/ISISexpl.jpg

[Lyman]

3hole

My issues are of design, and the use of trim in this case. The PF commanded a climb up without the assistance of TRIM. So either by good luck, or a/c design, the THS stayed put, and he relied on elevators only.

At the top, the design input TRIM after the STALL. That seems unwise. Commanded to be sure, but none the less, it responded.

Several of his inputs served to lower the NOSE, and the STALL sounded, and he pulled up again.

So, once STALLED, there would not only seem to be no need for all NU from the THS, but that it would be highly dangerous to leave it there.
With a continuous STALL warning, one would hope a 'smart' a/c would return the THS to neutral, and allow the Pilot to recover with elevators.

Since the speeds had returned, it becomes especially apparent (to the aircraft) full up THS is not a good thing? No alert, no lowering the THS,

I could easily be the thick one here, and not you.

[A33Zab]

It would appear that AB sent a FOT saying that the BUSS system should not be used above F/L 25.0 on 9th. September 2009.

Because UAS above FL 250 is considered transient. (caused by ICE)

One need to switch all 3 ADR P/B to OFF to activate BUSS scale.

http://i474.photobucket.com/albums/rr101/Zab999/BUSSScale.jpg

BUSS DESIGN
The BUSS is displayed on both PFDs. It is based on Angle Of Attack (AOA) information, and
enables to easily and safely fly the aircraft without any valid airspeed indication. The BUSS
enables to fly the aircraft in the entire flight domain while observing the design limits.
The BUSS scale is divided in three color-coded areas:
• In the green area, the aircraft has normal margin towards ultimate limits i.e. VMO/VFE for
upper limit and stall for lower limit
• In the upper and lower amber areas, the aircraft has limited margin towards ultimate limits.
Flying in the amber area is acceptable for a limited period of time but it should be avoided
• The red areas (FAST or SLOW) must be avoided.
The BUSS is tuned using the aircraft’s aerodynamic model with speed brakes retracted.
When the BUSS is active:
‐ The altitude indications are based on GPS data. Two amber dashes cover the last two
digits because the GPS altitude is less accurate than the barometric altitude
‐ The vertical speed indication is no longer available.
FLYING TECHNIQUE
The PF adjusts the pitch and thrust, to maintain the AOA in the green area of the speed
scale.
For approach, the flight crew should perform a stabilized approach. The flight crew should
change the aircraft configuration with wings level.

[Linktrained]

A33Zab

Thank you for the details of the ISIS.

Could you add the size in inches (or mm) across the horizon, on the A/H part, so that I ( and others interested ) can adjust the scale on a monitor to be able to appreciate the instrument properly, as it might be viewed by a PF or PNF.

[Lonewolf_50]

‐ The altitude indications are based on GPS data. Two amber dashes cover the last two digits because the GPS altitude is less accurate than the barometric altitude
‐ The vertical speed indication is no longer available.
Let me try to understand: static air data is deleted from the system? Is that what is going on in this mode?

[A33Zab]

http://i474.photobucket.com/albums/rr101/Zab999/ISISThales.jpg

[ChristiaanJ]

A33Zab,
Thanks for the pics and the info.
Was that the fit of the AF447 A330?

[GarageYears]

So, once STALLED, there would not only seem to be no need for all NU from the THS, but that it would be highly dangerous to leave it there.
With a continuous STALL warning, one would hope a 'smart' a/c would return the THS to neutral, and allow the Pilot to recover with elevators.

Since the speeds had returned, it becomes especially apparent (to the aircraft) full up THS is not a good thing? No alert, no lowering the THS,

So, you're advocating more automation? I'm not disagreeing here... I actually think all the information required to help the crew was available.

[A33Zab]

All ADR data is removed when ADRs are switched OFF.

With this BUSS option installed the AOA signal is fed into IR part of ADIRU,
IR part is available even with Air Data P/B off.
Don't know if more inputs are re-routed.

[RetiredF4]

Lonewolf_50

3hole, a few threads back, I think number 4, a gent who has taught in the A330 simulator related to us how some Unusual Attitude drills and recoveries required use of the trim wheel, manually, to get the pitch under control (in other than normal law, IIRC) in a suitable time frame. He wasn't sure if that was a "simism" or not.

Point?

There were or are some training drills that can be used to accustomize crews to using the trim wheel to control nose pitch in cases where the auto trim is either too slow or not helping redirect the elevators and THS where the pilot needs them. You could call the trim wheels a secondary flight control (Not sure if I am right about that) which means any rated pilot ought to know, like the flaps, when and how to use it.

And practice doing so.

Is it anywhere referenced to use it in this situation? I could only find the following references, where the mechanical trim is desscribed as standby system in case of autotrim not being availble.


THS MECHANICAL CONTROL
The mechanical pitch trim commmands are achieved by the handwheels in the
cockpit and transmitted to the input shaft of the THS actuator.
The THS mechanical control can be used :
- on ground, to trim the stabilizer when electrical control is not possible.
- in flight, as a standby system if automatic control (autotrim) is not available. (REF. LTTM)

The control wheels are used in case of major failure (Direct
Law or mechanical back-up) and have priority over any
other command. (REF: FCTM)

The purpose of the mechanical backup is to achieve all safety objectives in
MMEL dispatch condition: To manage a temporary and total electrical loss, the temporary loss of five fly-by-wire computers, the loss of both elevators, or the total loss of ailerons and spoilers.
It must be noted that it is very unlikely the mechanical backup will be used, due to the fly-by-wire architecture. For example, in case of electrical emergency configuration, or an all-engine flameout, alternate law remains available. In the unlikely event of such a failure, mechanical backup enables the PF to safely stabilize the aircraft, using the rudder and manual pitch trim, while reconfiguring the systems. (REF. FCTM)


When automatic pitch trim is no longer available, the PFD indicates this with an amber USE MAN PITCH TRIM" message below the FMA. (Ref. FCTM)


THS MECHANICAL CONTROL
General
The Trimmable Horizontal Stabilizer (THS) has a mechanical control system.
You can use the mechanical control system on the ground to trim the stabilizer when electrical control is not possible. The mechanical control system is also used in flight as a standby system if automatic control is not available.
The mechanical control transmits mechanical pitch-trim signals, put into the
hand wheels in the cockpit, to the input shaft of the THS actuator. (REF. LTTM)

We know now that it might have helped to defuse the situation, how could the AF447 crew know? By not knowing the above references?

[OK465]

BUSS looks suspiciously like an AOA indexer.

I 'spect it's possible to have an indexer without shutting off the all the ADR's also.

We had an indexer in the training 727 we flew in the 80's, as well as a separate full-time AOA gauge.

Back to the future.

[Lyman]

GY

Why wouldn't I be advocating for more automation? Or a better fit between what exists and the crew? Isn't that what this is all about?

If this airframe thinks enough of itself to protect the Rudder from inadvertent and harsh deflection, why would it not also want to protect itself from a TRIM trapped STALL?

If more than 8 degrees Rudder is not available at certain speeds, why would -13 PU be available? This is the "disconnect" I see from the outset of this thread, some logic here, some logic there, but something short of a continuous path in all directions.

It can be explained, but excuse me, pulling out the book and trying to get task specific recurrent in these conditions is insanity.

Knowledge plus experience equals wisdom. Substitute the word intuitive for wisdom, and there you go. Nothing that can occur on this aircraft can be permitted to be anything short of fluid, intuitive, and absolutely resistant to failure.

And that is the goal for when things are difficult, and doable; to say this argument is justifiable at all when problems of a familiar nature are involved is frankly outrageous.

It would be so nice to simply say PE. That is not any longer acceptable if one also expects to take credit for a failure resistant a/c when faults are so readily apparent?

We read here from guys who dream this stuff and live it. But there are real people here, and when some assume a posture of haughty self satisfaction that makes my fur hurt. The aircraft has holes, fill them.

all the best

[3holelover]

A33ZAB,
Many thanks for the details of the ISIS.... looks and sounds very much like the SAARU in the triple 7. ...an everything-you-need-PFD-in-a-self-contained-box. Pop one of those in your little ragwing and you're home free!

Lyman.... Thanks for clarifying... One further comment, re: "If this airframe thinks enough of itself to protect the Rudder from inadvertent and harsh deflection, why would it not also want to protect itself from a TRIM trapped STALL?"
Because it is NOT "trapped" at all and there is no risk of structural damage as there is with the rudder.

[infrequentflyer789]

Speed is decreasing and as speed is decreasing, they don't see it and what does the THS do on FBW machine?: It tries to maintain flight path, so as speed is decreasing to maintain flight path the autotrim goes up to the limit, incidence is growing up to maintain flight path.
On a Boeing, even in alternate law, you would have to trim as speed decrease to maintain flight path, it would alert you that something is wrong.

Autotrim will do exactly the same on a boeing if the automatics are trying to maintain a flight path (with speed decreasing). See Schipol 737 and similar accidents/incidents.

Not a FBW issue, not an AB issue.

[A33Zab]

Many thanks for the details of the ISIS.... looks and sounds very much like the SAARU in the triple 7. ...


Minor correction:

-ISFD (Integrated Standby Flight Display) is the B. naming for ISIS.
-SAARU (Secondary Attitude Air data Reference Unit) is the "back-up" ADIRU of the777.


an everything-you-need-PFD-in-a-self-contained-box. Pop one of those in your little ragwing and you're home free!



Exact! 1 standby instrument for whole fleet! A. thru Z.

[A33Zab]

For ISIS, this ISIS (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-4.html#post6645966) was installed.
The BUSS option was NOT installed.

[Lyman]

3hl

"No risk of damage, as with the Rudder?" Methinks the reason the THS was not deployed in the climb, and the elevators were not allowed to grab too much at a time had to do with protections.

The a/c at the bottom is not Damage?

The Rudder could disintegrate and part the VS, no crash. The THS, at full NU does not risk damage to the a/c? What if they had recovered aero flight, and the THS remained at -13 in a steep and high speed dive. No risk of damage? What a ride. ?

No inference of applicability, but look at this. The Rudder on AA587 tore the VS off (That was at <250 knots). If 447 had been in a recovery from Stall, the THS couldn't take the Tail off? I wouldn't place a bet.

[Machinbird]

http://i474.photobucket.com/albums/rr101/Zab999/BUSSScale.jpg
OK465BUSS looks suspiciously like an AOA indexer.I concur, Increasing AOA is toward the bottom. Oriented this way to make the display read in a manner similar to airspeed. Note: There are no numbers on the display.

Shouldn't the note at the bottom right say:
Red lower area corresponding to CAS < VLS

AF447 would have been at the bottom of the bottom red band while on its way down. Even three highly confused aviators would have understood their plight then.
Well, lets hope so.

[HarryMann]

Maybe the "whooler" shouldn't have been eliminated, but I'm not at all sure that would have made any difference here. It might have though...

I think a THS position trim clacker might well have alerted PNF to something going very awry in pitch much earlier and at least have glanced at the wheel or pointed the Captain to it immediately upon cockpit re-entry ?

.. even if PF himself was 'focused down a very narrow tunnel of perception' and would likely have been aurally blocking it out.

[airtren]

Speed is decreasing and as speed is decreasing, they don't see it and what does the THS do on FBW machine?: It tries to maintain flight path, so as speed is decreasing to maintain flight path the autotrim goes up to the limit, incidence is growing up to maintain flight path.

On a Boeing, even in alternate law, you would have to trim as speed decrease to maintain flight path, it would alert you that something is wrong.

Autotrim will do exactly the same on a boeing if the automatics are trying to maintain a flight path (with speed decreasing). See Schipol 737 and similar accidents/incidents.

Not a FBW issue, not an AB issue.
This is a drastic speed decrease, all the way down to a Stall condition, which is determined, and announced for 53 seconds by the "a/c's controls". Stall means fall !!!!

During these 53 seconds, of definite Stall Condition, determined by the a/c's computers, the THS went also under the a/c's computers control from -3 to -13 degree, which is MAX Nose Up, which is exactly the opposite of what in a Stall condition the pilot, or the computers should be commanding.

Personally, I would not consider this a FBW issue, in the sense of the FBW being a system in which the medium for transferring/passing bi-directionally control information is wire.

Is this really not an Issue??? if Boeing is doing to wrong thing, it is OK for AB to do the same?

[jcjeant]

Hi,

3holelover
Lyman.... Thanks for clarifying... One further comment, re: "If this airframe thinks enough of itself to protect the Rudder from inadvertent and harsh deflection, why would it not also want to protect itself from a TRIM trapped STALL?"
Because it is NOT "trapped" at all and there is no risk of structural damage as there is with the rudder. I dunno about structural damage for THS .. but this ended with a big structural damage (the entire plane and it's load)
Maybe it's to understand Lyman prose as:
Why the THS still full up .. when the plane know (seem's the pilots no) he is in a full stall
Plane know this is a full stall .. plane know this is a full up THS deflection
The two are contradictory
Why not automatically reset the THS in a better position (with a warning to pilots) like it's a limitation for rudder deflection in particular situations ?

[airtren]

So, once STALLED, there would not only seem to be no need for all NU from the THS, ....
With a continuous STALL warning, one would hope a 'smart' a/c would return the THS to neutral, and allow the Pilot to recover with elevators.

So, you're advocating more automation? I'm not disagreeing here... I actually think all the information required to help the crew was available.

It's not "more automation". It is rather "right automation"!!!:D

[USMCProbe]

All of the AOA information to avoid this crash are present in every modern Airbus aircraft, as standard equipment. In addition to the stall warning and the VLS/Alpha Prot, Alpha max scale on the left side of the PFD, there is an even better, direct measure of AOA, even though it does not come from the AOA installed on the aircraft (3 on the 320 series).

The problem is, no one is trained to use it.

What is AOA? It is the angular difference between the relative wind and the longitudinal axis of the aircraft (chord of the wing, more specifically). In lay-mans terms (I am one), it is the difference between where the airplane is pointed, and where it is going, about the lateral axis.

Call up the "bird". With the wings level, the AOA is the vertical distance from the pitch bars (where the aircraft is pointed), and the "bird" (where the aircraft is actually going). Non-wings level, you visually drop a perpendicular from the plane of the pitch bars to the bird. That distance, measured on the PFD, is your AOA, and would have prevented this crash.

I know some engineer (I am one also) will say that this AOA measurement is not "perfect", but neither is an AOA gauge. Not even close. Just as IVSI is more useful than a pure VSI, this IRU based measurement is actually more user friendly to the pilot, more reliable, and "cleaner". And it is EXACTLY where the pilot is staring to control the aircraft, there is no other "scan" to learn.

On a 320 series, cruise is usually 2-3 degrees up, green dot, S, and F speed are 7-8, Stall warning (in alternate law in the sim) comes on at 10-12 degrees with the stall happening a bit above that. In level flight this is the "pitch attitude". In non-level flight, this is the perpendicular distance from the pitch bars to the bird. Icing doesn't affect the IRU's and there are 3 of them on the 320.

We were just not trained to use this.

[Lyman]

The pilot(s) did call up the bird, FPV, but it was not working. That from ACARS, I think.

[3holelover]

It seems to me....
...that many of you are looking for all sorts of things that either automation or aircraft systems of one sort or another could have done/be done in future to help in this situation, while forgetting that flying is an inherently dangerous undertaking that has, for eons now, been made rather routine by simply training pilots how to handle their machines.

Yes, you could automate the hell out of the thing, but at some point, somehow, all that automation could easily be counter productive.... Can't you imagine a situation, for instance, where a stab that returns to neutral automatically, might be a bad thing? Maybe it's best to leave the aircraft handling to pilots?
Maybe it's better to train pilots to fly, rather than manage computers?

This aircraft didn't fail. One system alone had what should have been a non earth shattering, temporary fault, due to ice crystals, but then it's pilot(s) failed.

Yes, as I said earlier, I agree there are probably some tweaks that might make sense, but I don't think anyone should get carried away in burying human failure with more complexity and more tacked-on, automated garbage that will wind up fouling somebody else down the road.

As yet, nobody wants to fly in a fully automated, pilotless aircraft. All we need, to get folks to their destination safely, in any serviceable aircraft, is a pilot who knows his machine, and knows how to fly it. That simple formula has been proven to be quite safe. What was missing here wasn't more junk, it was human knowledge and skill. Airmanship.

Does anyone here really believe an aircraft can be made that won't stall? My limited knowledge of aerodynamics and physics tells me, with current aircraft design, that's not possible. So pilots had better know how to recognize and correct for a stall. The surprise to me here, has been learning that hasn't been universally the case. That, imho, needs immediate correction!

[airtren]

It seems to me....
...that many of you are looking for all sorts of things that either automation or aircraft systems of one sort or another could have done/be done in future to help in this situation, while forgetting that flying is an inherently dangerous undertaking that has, for eons now, been made rather routine by simply training pilots how to handle their machines.

Flying is inherently dangerous, but how a/c's got better and better from their original inception, was also from understanding the shortcomings and improving or eliminating them. Combining a better training with an improved a/c is a better combination than just training.

Yes, you could automate the hell out of the thing, but at some point, somehow, all that automation could easily be counter productive.... Can't you imagine a situation, for instance, where a stab that returns to neutral automatically, might be a bad thing? Maybe it's best to leave the aircraft handling to pilots?
Maybe it's better to train pilots to fly, rather than manage computers?
The current model is only going to progress on even further automation.
What is needed, is to eliminate or minimize the shortcomings in the design, which will reduce the risks of manual flying..

This aircraft didn't fail. One system alone had what should have been a non earth shattering, temporary fault, due to ice crystals, but then it's pilot(s) failed.
Certain elements of the current design had their own contribution to the failure of the pilots. Improving those elements will only make the design, the a/c better, and the flying safer.


Yes, as I said earlier, I agree there are probably some tweaks that might make sense, but I don't think anyone should get carried away in burying human failure with more complexity and more tacked-on, automated garbage that will wind up fouling somebody else down the road.
don't see it as more complexity, but exactly the opposite less complexity. I think much of the current complexity is so - complex - because it is counter-intuitive. Making it more intuitive is making it simpler and easier for the pilots to quickly understand. From the information regarding the active stick position/motion shared directly among the pilots, to a more logical Stall Warning, to an Autotrim that locks, making sense in a Stall condition... etc.


As yet, nobody wants to fly in a fully automated, pilotless aircraft. All we need, to get folks to their destination safely, in any serviceable aircraft, is a pilot who knows his machine, and knows how to fly it. That simple formula has been proven to be quite safe. What was missing here wasn't more junk, it was human knowledge and skill. Airmanship.
We've been at this since the first passenger airplane: we have continuously complemented airplane shortcomings with pilot talent and training. But... in the same time we have improved a/c's after each accident that made us understand better its shortcomings.


Does anyone here really believe an aircraft can be made that won't stall? My limited knowledge of aerodynamics and physics tells me, with current aircraft design, that's not possible. This a/c showed a remarkable stability and extraordinary structural integrity during the Stall. That is exceptional.

It would be nice to be able to make it so it would not stall. But I don't see that as an attainable goal.

However, it is achievable to improve the logic, the algorithms, the software subsystem, so that remarkable stability, and extraordinary structural integrity is not wasted if in a stall, by making the stall avoidance better, and stall recovery easier.

[Machinbird]

3Hole
Don't you see a problem with an aircraft with an automatic trim system that trims silently? For years we have had alerting devices on THS trim movement (clackers & whoolers) to let you know the thing is moving. Then comes the new guy on the block with a "better" idea. (Airbus)

Couldn't they have seen the trim moving in their peripheral vision? Maybe, maybe not. There were serious problems in the cockpit that night that demanded attention.

If someone in the cockpit had noted that trim running up, they could have stopped it just like you would in a 707. Grab the trim wheel. Without all that nose up trim, the PF would have had serious trouble holding the nose up against the natural aerodynamic tendencies of the basic aircraft to drop its nose as it slowed. The PNF could have largely overridden the inept handling of the PF if it had occurred to him.

But no one saw the trim move. No one heard it move. If they had thought about it, they should have anticipated it moving, but they had other problems and it bit them.

We have now lost two Airbus aircraft due to lack of crew awareness of trim position and (other problems). Fixing the trim system to provide better motion alerting should be easy. It has plenty of precedents and it is not rocket sciences.

[3holelover]

Yessir Machinebird. I agree. That'd be a worthwhile "tweak".

[airtren]

Machinbird,

It is clear that the silent trim was not helpful in that stressful situation, and you've eloquently explained that.

But I am going further.

The trim should have stopped, or should have never happen during the Stall.

The a/c computers determined that the a/c was in a Stall, and sounded the alarm for 53 seconds.

During those very 53 seconds, the autotrim moved silently the THS to its extreme position of -13 degrees of max NU.

It is clear that there was a logical disconnect between the Stall and Autotrim. Had there been a logical connection, with a locking of the THS in Neutral, it would have certainly helped the pilots, the recovery from Stall.

During the rest of the descent, the Stall Warning sounded another 8 times, and the THS has remained in its max NU during all those 8 Stall Warnings.

Should the THS have returned to Neutral at the first Stall Warning out of those 8, the recovery would have been helped.

3Hole
Don't you see a problem with an aircraft with an automatic trim system that trims silently? For years we have had alerting devices on THS trim movement (clackers & whoolers) to let you know the thing is moving. Then comes the new guy on the block with a "better" idea. (Airbus)

Couldn't they have seen the trim moving in their peripheral vision? Maybe, maybe not. There were serious problems in the cockpit that night that demanded attention.

If someone in the cockpit had noted that trim running up, they could have stopped it just like you would in a 707. Grab the trim wheel. Without all that nose up trim, the PF would have had serious trouble holding the nose up against the natural aerodynamic tendencies of the basic aircraft to drop its nose as it slowed. The PNF could have largely overridden the inept handling of the PF if it had occurred to him.

But no one saw the trim move. No one heard it move. If they had thought about it, they should have anticipated it moving, but they had other problems and it bit them.

We have now lost two Airbus aircraft due to lack of crew awareness of trim position and (other problems). Fixing the trim system to provide better motion alerting should be easy. It has plenty of precedents and it is not rocket sciences.

[jcjeant]

Hi,

This aircraft didn't fail. One system alone had what should have been a non earth shattering, temporary fault, due to ice crystals, but then it's pilot(s) failed. I will repeat this X1000 if needed:
The Pitot tube was not faulty
As many here like write about the plane ...
The pitot tube operated normally .. as required by its specifications and certifications
The problem is that the measuring instrument has been used outside his operating range
Ice crystals is not a area for use Pitot tube .. it's not in the certifications
As the plane went well until he was out of its flight envelop
As the pilots went well .. until they were (or put their self) in a situation out of their skill

[Machinbird]

Airtren,
We are not too far apart in viewpoint. The nose up trim should have stopped running well before reaching alpha stall in any case.

[jcjeant]

Hi,

Fixing the trim system to provide better motion alerting should be easy. It has plenty of precedents and it is not rocket sciences.


Pending a resolution of Airbus engineers and certification in 10 years .. pilots can easily solve the problem of silent motion trim
No rocket sciences indeed ...

http://i.imgur.com/ZCNhm.jpg

[GarageYears]

It is clear that there was a logical disconnect between the Stall and Autotrim. Had there been a logical connection, with a locking of the THS in Neutral, it would have certainly helped the pilots, the recovery from Stall.

I think you have to be careful what you ask for.

- Making the trim operation audible (this seems like a no brainer). Relying on the visual cues seems over optimistic, particularly when things have already gone pear-shaped.

- Auto-neutralizing trim state in stall warning. Unless I am misunderstanding things here, the trim was not what stood this aircraft on it's tail, it was the PF elevator input. My understanding is that the elevators alone can override the THS. Would neutralizing the THS have been the key here to save the aircraft? I don't think so at all. The PF (and PNF+Captain) at no point acknowledged the STALL. In fact the PF was demanding NU pitch for most of the time - he got what he asked for. Would the PF have NOT demanded NU if the THS has not followed his demands... or had untrimmed? Doubtful.

I do believe a better more explicit YOU ARE STALLED alerting system can be implemented. The aircraft knows enough from AoA, Vertical Speed and perhaps Ground Speed to be able to state quite happily "You are stalled and plummeting to certain doom - how about reducing AoA with a bit of nose down old chap...". OK may be that's a bit glib, but something beyond the existing Stall Warning audible ---> "AIRCRAFT STALLED! AIRCRAFT STALLED! REDUCE AoA!"

[lomapaseo]

jcjeant

I will repeat this X1000 if needed:
The Pitot tube was not faulty
As many here like write about the plane ...
The pitot tube operated normally .. as required by its specifications and certifications
The problem is that the measuring instrument has been used outside his operating range
Ice crystals is not a area for use Pitot tube .. it's not in the certifications
As the plane went well until he was out of its flight envelop
As the pilots went well .. until they were (or put their self) in a situation out of their skill


I believe that you are too quick to defend the pitot tube as not being at fault. The earlier poster had specifically stated "pitot system"

It was the speed measuring system at fault (the pitot was only a part of it). Of course you are correct that such a system fault was presumed at aircraft certification..

At this point the assumption is that a combination of systems or unlikely events probably caused the accident. The issue now is what needs to be fixed to prevent another accident.

It is not enough to change the pitot tube alone when clearly a system fault was not accomodated to continued safe flight and landing.

[rudderrudderrat]

Hi Garage Years,

I agree. It is pointless adding more audible warnings to already saturated pilots.
Altitude Alert, Multiple Ecam Dings, Autopilot Disconnect, Thrust Lock Dings etc. + "Stall Stall".

A different warning / alert is required to elevate it to the highest attention getter - like the discrete stall warning vibrations through the control column on other aircraft. It simulates the natural buffet onset. There is no other warning like it.
The audible "Stall Stall" seems to have been filtered out of all 3 crew members sensory perception - no one commented on it.

[airtren]

....

- Auto-neutralizing trim state in stall warning. Unless I am misunderstanding things here, the trim was not what stood this aircraft on it's tail, it was the PF elevator input. My understanding is that the elevators alone can override the THS. Would neutralizing the THS have been the key here to save the aircraft? I don't think so at all.

The "THS at max NU" was a major immediate reducing - if not neutralizing - factor of the effectiveness of the "Elevators ND (even if at max ND)".

The "THS Neutral + Elevators ND" would have allowed an immediate more effective Elevators ND, compared with the "THS max NU + Elevators ND".


I do believe a better more explicit YOU ARE STALLED alerting system can be implemented. ...

On the Stall Warning, as it can be seen from my posts on the topic, we're on the same page.

[Lonewolf_50]

Retired F4

As caveated by our contributor, the scenario might not have been similar enough to have triggered a "this is one of those situations" responses from the crew.

What I was trying to get at (damping and pitch rate limiting functions considered) is that I think the pitch rates would be a little faster (you'd have to wait less time) if you moved the THS with the wheels (even when auto trim is functioning, which it should in Alt 2) rather than wait for the elevator THS linkage to induce THS nose down (THS actually up a bit) commands.

But if one doesn't train it, one probably doesn't resort to such actionns when things get hairy in the cockpit.

[deSitter]

How hard would it be to just post a little animated graphic of the fin/stab configuration? Something that can be instantly visually absorbed? "Wow that can't be right.."

[jcjeant]

Hi,

lomapaseao
I believe that you are too quick to defend the pitot tube as not being at fault. The earlier poster had specifically stated "pitot system"

It was the speed measuring system at fault (the pitot was only a part of it). Of course you are correct that such a system fault was presumed at aircraft certification..The Pitot tube is not only a part of it .. it's the most important part of it
Remove from the plane the Pitot tube and you have no more speed measuring system at all
The AF447 case proven this.
The best solution at today date is to use the Pitot tube in his domain of certification
So .. dont fly in ice crystal area ..
If not able to detect ice crystal .. don't fly in the areas when this is the possibility to meet them .. forget fuel sparing .. fly safe !

[RetiredF4]

Lonewolf_50

Retired F4

As caveated by our contributor, the scenario might not have been similar enough to have triggered a "this is one of those situations" responses from the crew.

What I was trying to get at (damping and pitch rate limiting functions considered) is that I think the pitch rates would be a little faster (you'd have to wait less time) if you moved the THS with the wheels (even when auto trim is functioning, which it should in Alt 2) rather than wait for the elevator THS linkage to induce THS nose down (THS actually up a bit) commands.

But if one doesn't train it, one probably doesn't resort to such actionns when things get hairy in the cockpit.


I´m fully with you at that, but:
I wanted to make it clear, that i couldn´t find a reference in the valid publications for aircrew (FCOM, FTCOM) and maintenance (LTTM), which even would give a hint to use manual trim in such cases.

So how should the crew know about it?
Why should it have been trained?
Did anybody knew it before AF447?
On what basis would it be trained now?

We had the manual trim discussion before on the older thread, but there everybody asssumed direct law, where only manual trim is available.

We even dont know, how the system will behave after manual trim has been applied despite autotrim still active.

(Ref. LTTM)

Operation/Control and Indicating
When you turn the handwheel installed in the cockpit center pedestal the chain and cable loop move the input shaft. The input shaft moves the mechanical servoloop mechanism through an override mechanism. The override mechanism, which is installed adjacent to the Pitch Trim Actuator (PTA), makes sure that the mechanical control cancels the electrical control.


Operation of the Override Mechanism

Operation of the Mechanical Control

The input shaft turns.

A cam in the shape of a ’V’ turns.

The cam moves a roller which disconnects the internal microteeth from the
output shaft of the PTA.

At the same time a piston is pushed to operate the three override mechanism microswitches.

The mechanical control now comes on before the electrical control.

Release of the Mechanical Control


The input shaft stops.

The cam in the PTA is released.

The internal springs connect the internal microteeth to the output shaft of
the PTA.

At the same time the piston moves back from the microswitches.

The mechanism is now set in the electrical control position.



MECHANICAL INPUT SHAFT
The Mechanical input shaft connects the mechanical control loop and theTHS
actuator. The shaft has a flange which is connected to the mechanical input of the THS actuator. The other end of the shaft has a sprocket. The mechanical control-loop chain is installed on the sprocket

What happens after release (which i interpret as not turning the wheel any more)? I couldn´t find anything in the manuals.

Point is: We know now, that trimming the THS manually would have helped to bring the nose down. But we dont know consequences in other circumstances, as there seems to be no documentation concerning such a usage. So how can one train for it without documentation?

[deSitter]

NASA had an orbital configuration display software one would often see during slow spots in Shuttle mission coverage. One glance at this screen and the configuration of the craft relative to the Earth was instantly obvious. In the open areas of the display one found all the necessary data to have a precise description of the Shuttle's configuration.

Why cannot a modern aircraft have such a display for itself?

[Lyman]

Not one proposition offered to mitigate the THS fatal flaw here was NOT looked at by Airbus.

One thinks one can solve a problem that one assumes to have been OVERLOOKED? By AIRBUS?

3hl : perhaps the solution to a "go to your room" command to THS in other than NORMAL LAW is not necessary afterall.

Airbus, in their wisdom created AL2 with ROLL DIRECT. PITCH NOT DIRECT.

They still wanted the a/c to have some "say so"

The pilots can be trusted with ailerons and spoilers, but not the TAIL.

Initially, when the THS did NOT contribute to the zoom climb, it looks like a good idea.

Let the elevators nudge the a/c into a crazy climb with the Computers help? YES. The climb is a pattern of UP nibbles and retreats, in thrall to the PF's STICK. With each nudge to a higher AoA, the computer forgot the last one, and the zoom climb resulted. Then at the top, it forgets its last nonsense, and tries to maintain a load.

Going up, maintain load, at the top maintain load. The PF is not the only one who appears to have lost his "mind".

So the THS was not "Jammed"? Baloney. Two powerful motors were holding it against the Top curb. It was free to move only upon command, a command of NOSE DOWN. It recieved plenty of these, but the computer disallowed them v/v THS.

Pilot responsibility? Of course. AIRBUS? You make the call.

If elevators can (and should be) used to recover, or maneuver, or maintain, etc. Whose idea was it to throw in a silent partner?

Flying is NOT inherently Dangerous, that is a bromide, and demonstrable for those of us who hang on to the hero pilot meme.

It is inherently SAFE, as the statistics PROVE.

[airtren]

jcjeant,

If I read again your previous post, I think it is quite a bit about semantics.

It is easy to agree with your saying that the "pitot tube was not faulty", as I take that as the "pitot tubes were not defective", as the pitot tubes functioned normally in different conditions than those when they failed.

But even if they were not faulty, they did fail, as a device can fail, without being defective, when in an abnormal operation conditions.

But, IMO, you're raising a high threshold for yourself in saying: "The pitot tube operated normally .. as required by its specifications and certifications". The "required" implies IMO that the language in the specs/certifications is strong, some wording that is equivalent to the combination /MUST/MUST NOT/, i.e., mandate the freezing in the conditions of their failure, and mandate the providing of inaccurate information in the condition of their failure.

Are you sure about meeting that threshold? ...

"
The Pitot tube was not faulty
...
The pitot tube operated normally .. as required by its specifications and certifications
The problem is that the measuring instrument has been used outside his operating range
Ice crystals is not a area for use Pitot tube .. it's not in the certifications
As the plane went well until he was out of its flight envelope". Hi,

lomapaseao
The Pitot tube is not only a part of it .. it's the most important part of it
Remove from the plane the Pitot tube and you have no more speed measuring system at all
The AF447 case proven this.
The best solution at today date is to use the Pitot tube in his domain of certification
So .. dont fly in ice crystal area ..
If not able to detect ice crystal .. don't fly in the areas when this is the possibility to meet them .. forget fuel sparing .. fly safe !

[OK465]

From post #101:


Call up the "bird". With the wings level, the AOA is the vertical distance from the pitch bars (where the aircraft is pointed), and the "bird" (where the aircraft is actually going). Non-wings level, you visually drop a perpendicular from the plane of the pitch bars to the bird. That distance, measured on the PFD, is your AOA, and would have prevented this crash.


@USMCProbe:

There's one small problem with this. At 16 degrees pitch attitude and AOA in excess of 35 degrees, do you know where the "bird" is on the PFD?

(Any 'stall prevent' use of the "bird" would have required monitoring and honoring FPV ("bird") thru the "ballistic" portion of the "zoom", the caveat of course, it must be selected and available.)

edit: BTW there are a lot of pilots trained to use FPV.

[jcjeant]

Hi,

But, IMO, you're raising a high threshold for yourself in saying: "The pitot tube operated normally .. as required by its specifications and certifications". The "required" implies IMO that the language in the specs/certifications is strong, some wording that is equivalent to the combination /MUST/MUST NOT/, i.e., mandate the freezing in the conditions of their failure, and mandate the providing of inaccurate information in the condition of their failure.The language used in certification .. the words used must be accurate and not misleading and should not allow an interpretation that could afford not to keep strictly to these specifications
If this is not the case these certifications are useless and therefore the classification societies and certifications societies (and regulators) must be submitted to an external audit to detect problems .. skills or organizational
Trials and penalties can be at the corner .. as those specifications concern not only technical matters but also human lifes.

[airtren]

As a clarification, I think, like you, I expect the specs/certs to be non-ambiguous in defining the operational conditions, and functions within those conditions.

Without the specs/certs in front of me, I don't know if that non-ambiguous language extends to the range of conditions which is outside the operational conditions, as that's outside the scope. Therefore, I asked my question, as the possible ambiguity for the outside operational conditions range creates the difficulty of meeting the threshold, which I think you've set for yourself, with the wording you've used.

Hi,

The language used in certification .. the words used must be accurate and not misleading and should not allow an interpretation that could afford not to keep strictly to these specifications
If this is not the case these certifications are useless and therefore the classification societies and certifications societies (and regulators) must be submitted to an external audit to detect problems .. skills or organizational
Trials and penalties can be at the corner .. as those specifications concern not only technical matters but also human lifes.

[HarryMann]

Ok, enough of this THS speculation.. I (and others) mooted extreme concerns over this (misuse case) style of behaviour hundreds of posts ago... a few others didn't agree, or thought we were grasping at straws - fair enough.

Hopefully, and looking at BEA wording, it will not be ignored and pushed to one side - everything is relevant, even irrevelances (in that they are tagged irrelevant!)

Now, we can look a bit further away... ?

I can certainly see that calls for some 'neutering or neutralising' THS logic in 'apparent' stall situations raises eyebrows, mine as well.

What goes right back to basic aircraft certification requirements in the 60's, is the stick-pusher. Indeed, RudderRudderRat just suggsted a stick shaker mode would be somethingt that makes some sense

So far, AFAIK, certification requirements mandated combined shaker and stick pushers for only those aircraft with deep (e.g. potentially irrecoverable) characteristics.

I would think AoA measurement has become easily reliable enough for this to become a safe option for all airliners above a certain size... the prelimary vibration itself could quite possibly have been enough to shake PF out of what may have tunnel vision.
THS activity and detected position would automatically have to be become part of stick shaker activity.

Airtren.. damn good summing up above :D

[lomapaseo]

The Pitot tube is not only a part of it .. it's the most important part of it
Remove from the plane the Pitot tube and you have no more speed measuring system at all
The AF447 case proven this.
The best solution at today date is to use the Pitot tube in his domain of certification
So .. dont fly in ice crystal area ..
If not able to detect ice crystal .. don't fly in the areas when this is the possibility to meet them .. forget fuel sparing .. fly safe !

If we intend to hold somebody accountable to fix something than semantics is damn important lest they end up saying "it ain't my job" Improving the Pitot tube or avoiding weather entirely (because we can't readily see the stuff) will not prevent the next accident when this system fails in a different way (flocking birds, volcanic ash or ice sheds from in front of the pitots or even a loss of vision or whatever powers the instruments that displays the speed) That is what I mean by a system failure.

The system failed ....l so what? To fix the problem we need to examine not just this one specific failure condition but to revisit the assumptions in the basic certification (which assumes this system would fail) and address every one of the assumptions that was violated.l

I have no respect for the assignment of cause "but-for the icing of the pitot" that is typical of the legal tort system when it comes to adressing flight safety shortfalls in complex system interactions.

IMO this problem is not going away with just the final accident report summary or all the rhetoric of PPRune "I told you so" until wide ranging recommendations are adopted..

[Linktrained]

Certification (s)

Of course certification must keep up with increasing experience, whether for speed or pitch information.

Under the Regulations of sixty years ago any aircraft ( A330 ?) would be restricted to flights not exceeding 1000 nm over water or 1500nm overland. Longer flights required the carriage and use of a Flight Navigator, who would need a sextant and a means of checking the compass (Astrocompass ?) and a Drift sight. (Please don't suggest an astrodome on a pressurised aircraft - I think that someone lost their Navigator through one, then !)

There was a list published of some 30 odd UAS incidents which had not had MAJOR problems. Someone ought to know why. (NWA had ice crystals and were 25 miles away from a Cb.)

Historically many/ most aircraft accidents seem to have come from the convergence of several (often three) factors, each relatively minor or individually surmountable. Each factor, when indentified or even suspected ought to be remedied, so that THAT factor does not happen again.

[VGCM66]

If you are pilot, you better know what you are getting into. Nobody is building any airplanes that flies by itself. If you as a pilot do not comprehend the meaning and the peril of what a stall is and it's dreadful implication and possible fatal consequence, you have no business being in a cockpit and even less being in charge.
I am no pilot but even I know that when flying the main objective is not to FALL as in hitting Mother Earth FALL. By the time Altitude became meaningful again in that AF447 cockpit, it was too late.

THS did what it was supposed to do. If the PF wouldn't have applied NU for more than 3/4 of the 4 minutes (give or take)of the incident opening and closing the THS would have never moved at all.
ND would have probably brought the THS down if application of the SS would have lasted long enough.

The pitots have already been dealt with by the industry which incidentally just acted in flight AF447 as a casual instigators of life eternal pop test quizzes.

A captain knowing that bad weather was approaching ahead, decided it was a good time to leave the cabin. Say what?

The PNF (First Officer and probably future candidate for Captain) heard at 02 h 10 min 10,4 : SV : “Stall, stall” (without cricket). And he said: What is that? EXCUSE ME? "You ain't a passenger Mr.! You should and ought to know what STALL, STALL means preferably waaaaaaaaaaaaaaaaaayyyy before that tragic night. On top, later he can be heard desperately calling for the captain to come back to the cockpit. Don't know, but I think he was supposed to have acted as replacement captain not a captain caller. A pilot might need schooling to learn how to recover from a Stall of any kind but he/she does not need a whole schooling to learn that a Stall might get you kill along with all your passengers. That's my take.

And then there is the young PF. What can we say that can bring some meaning, reason or just some measurement of mitigation? That he was young, maybe? Not even.

A flaw was finally found and subsequently corrected in the Comet. But no R-E-D-E-S-I-N-G was ever made except for the upgrade to the Comet 4. They serve their time proud and tall afterwards. There are almost 1,000 A330 already made and in service. Any major redesign will be just done on newer future versions of Wide Body Twin Engine Models.

[Lonewolf_50]

Retired, I believe we did discuss this a couple of threads ago, and the concensus was that in Alt Law

If you use the trim wheels to move the THS, it will over ride the auto trim. Once moved, and then left alone, in Alt Law auto trim will again move it in response to the usual inputs.

Is this trained in the Sim?

Again, the point was raised in re some Unusual Attitude training scenarios, which isn't the same as an Airspeed Unreliable scenario.

I think we are in violent agreement, more or less. :cool:

[Lyman]

A casual observation. If some 'blend' of manual/auto commands (THS) are mandated, do we have an alert then, in/out? Prolly not. "Who's on first?"

In a DIRECT LAW situation, only manual. So in between, and true to "gradual" and "graceful" degradation, a fuzzy mush of "What are WE doing now?" The Waltz, or the Lambada?

With the brain lock some see here, is that at all fair?

Just because it's too much for me to imagine, I'm sure it's just the thing.

More prompts, more recorded music, more Synthetic verbalie. Soon, a Choir, then a symphony?

When run out of ideas, make it merely more complex? There is a gaping hole in there between Cruise and recovery from UAS. Those who downplay it should be chastened by now? Perhaps not, given the onset of the academy awards for those sold on the platform without reservation.

[RetiredF4]

Please,

if you are new to this thread and want to contribute something useful out of your expierience as an engineer or any other profession you are familiar with, it would be appropriate to read the nearly 1.000 pages filled about this flight.
That would tune you in the loop of the discussion.

There had been failures, there had been mistakes, and there are things which can be improved to reduce the probability of similar accidents.
The big question to all these matters is "why"?

Your ranting does not contribute to this task.

[Clandestino]

If the a/c can get into a STALL, it should be able to get out the STALL with Elevators.
There's no such certification requirement as it would be impractical to design aeroplane this way - please refer to DP Davies' masterpiece, page 35 onwards. Also it is entirely possible to certify passenger transport aeroplane that once stalled, cannot be unstalled by anything except improbable amount of luck, produce it in hundreds and have fly around for decades in statistically acceptable safety. It's not A330, if you wonder. Again, DP Davies' work is painfully relevant. Surprisingly to some, not me.

why would -13 PU be availableTo meet certification requirements of stability and maneuverability at diferent speeds, altitudes, weights, configurations and CG positions.

Why the THS still full up .. when the plane know (seem's the pilots no) he is in a full stall With all three ADIRUs losing speed signal, aeroplane is so confused as to know nothing. Mind you, I 've used "confused" and "know" figuratively. No computer installed on A330 is intelligent therefore it really doesn't know anything and it cannot be possibly puzzled. It works within limits of its instruction code.

THS went up to trim aeroplane into AoA it could not possibly sustain. Triggering condition was sidestick pitch command.

It is clear that the silent trim was not helpful in that stressful situationThere's good chance that even if there were trim in motion aural warning, it would be overridden by stall warning. OTOH, simple mechanical device like bicycle bell would be heard all the time during normal ops and probably wouldn't be noticed after a couple hundred of hours in flightdeck.

My understanding is that the elevators alone can override the THS.They can't. DP Davies explains why, page 35 and on.

It is not enough to change the pitot tube alone when clearly a system fault was not accomodated to continued safe flight and landing. There were at least six cases of total airspeed indication loss due high altitude icing before 447 on AF 330/340 fleet alone and allegedly there were about 30 cases worldwide. That every affected flight, bar one, continued and ended safely somewhat contradicts the notion of systemic failure.

So .. dont fly in ice crystal area ..
If not able to detect ice crystal .. don't fly in the areas when this is the possibility to meet them .. forget fuel sparing .. fly safe !
Big chunk of my flying is done through clouds made of ice crystals and I have never had my pitots frozen. Not every ice crystal will stick to your probes or compressors and ideas when and where can we encounter the dangerous type are currently extremely vague. If we were to avoid every CB plume in the sky, soon we'd measure our deviations in hundreds of miles and schedules would become a joke.

It seems to me....
...that many of you are looking for all sorts of things that either automation or aircraft systems of one sort or another could have done/be done in future to help in this situation, while forgetting that flying is an inherently dangerous undertaking that has, for eons now, been made rather routine by simply training pilots how to handle their machines.
You are not mistaken. Not a little bit.

Flying is NOT inherently Dangerous, that is a bromide, and demonstrable for those of us who hang on to the hero pilot meme.

It is inherently SAFE, as the statistics PROVE.

Utter lie, no, not even a lie, it is pure BS. Flying is fantastically dangerous. Statistics is hard science and can't prove inherent safety of anything at all. What your numerology that you have the cheek to call statistics fails to appreciate is how everyone in aviation needs to work his or her butt off to make potentially very lethal activity into something that the ignorant considers safe based solely on outcomes.

If only sheer ugliness of twisted bits of metal that used to be an aeroplane, now littering the seafloor or some mountainside, breaks up your dream of flight being safe because it's not dangerous in first place, don't come to PPRuNe for lullaby to put you back to sleep.

[Lonewolf_50]

bear
A casual observation. If some 'blend' of manual/auto commands (THS) are mandated, do we have an alert then, in/out? Prolly not. "Who's on first?"

In a DIRECT LAW situation, only manual. So in between, and true to "gradual" and "graceful" degradation, a fuzzy mush of "What are WE doing now?" The Waltz, or the Lambada?

One of the reasons in getting a type rating, and qualified on a given aircraft model, is to study it, fly it and by so doing learn how it works, and what makes it work well enough to safely operate it.

Each aircraft will have some different characteristics than others. The art of design continually evolves, so it should not surprise us.

What most professional pilots do is, rather than grouse about whether a particular combination of features pleases or doesn't, is learn how it works and how to make it fly well.

The objective of a professional training program is to make the above happen.

[Linktrained]

RetiredF4

Just like Orville and Wilbur, many of us have started as Newbys at some stage. Some of us may have learned ( and understood ) everything in a single presentation. Others may be helped by a further explanation or example, even when the initial version may prove to have been in need of even wiser thoughts, often from people like yourself.

[Mr Optimistic]

There were 3 trained pilots in this a/c. Although you can argue that the man machine interface had its failings (stall warner behaviour, lack of explicit AoA, failure to scream the THS was at the limit of its authority), the concern about pilotless aircraft is that human beings are adaptable and flexible and cannot be programmed out. This accident makes you wonder about the value of human beings versus machines: did they deliver their value here ?

[Lyman]

Sounds good, lonewolf.

#1 It didn't work.

#2 I have the distinct impression some qualifieds here are not in synch with the A330, or with your anlaysis, which I personally find impeccable.

#3 The difficulty with this Discussion is its free flow, and that is fine.

My opinions are based on my experience, and knowledge. As such, they mean probably nothing to anyone else. So what. Tell me, as a helicopter pilot, your impressions of having a combination manual/auto Trim on your aircraft. In Alternate Law, a regime virtually unflown in these circumstances by most if not all of the pilots currently operating the A300xx. On a moment's notice. No Sim, No Button. No ReDo.

It is so easy to focus on one thing, and that, as a pilot, can be deadly, hence your excellent posts on SA.

You scold the pilot group with innuendo that "Learn the Airplane"? is what's missing, the platform shows no unscrutables?

Or was it meant for me? Either way, it is not productive, and unlike your proven and excellent understanding.

Thousands of people have DIED to make flying Safe. It IS safe. Inherently Dangerous sounds like a line from a B movie. Driving is more dangerous, statistically, and in personal experience. Not all who fly are dumb.

I watched once at a test flight prior Launch. It was a wicked new and complex machine. The Pilot was confident, no, arrogant. fifteen minutes later he was burned beyond recognition, along with his a/c.

That isn't danger, that is insanity. The bottom line is the argument at all, here. Entrenched on the one side, over confident and arrogant on the other. A finding of any kind is pounced on, and morphed into "evidence." For what? EGO, and PRIDE. To my way of thinking, in furtherance of making flying LESS SAFE.

Personalyzing the debate is irritating. There is great stuff here along with the dross. Me Dross? Obviously, according to some, even most. Perhaps in my own eyes.

Mark my words, Aviation is careering into the weeds in front of us, and absent some patient direction from a source heretofore unknown, it will succeed in its quest for BBR.

[CONF iture]

THS went up to trim aeroplane into AoA it could not possibly sustain. Triggering condition was sidestick pitch command.
How stall warning didn't supercede that condition ... ?
Curious to see if the BEA will be curious about it ...

[jcjeant]

Hi,

Clandestino:
Big chunk of my flying is done through clouds made of ice crystals and I have never had my pitots frozen. Not every ice crystal will stick to your probes or compressors and ideas when and where can we encounter the dangerous type are currently extremely vague. If we were to avoid every CB plume in the sky, soon we'd measure our deviations in hundreds of miles and schedules would become a joke.So certainly with the insight of the AF447 accident you must think that the pilots were in very bad luck for have the 3 Pitot tubes frozen dead ..

If we were to avoid every CB plume in the sky, soon we'd measure our deviations in hundreds of miles and schedules would become a joke.

And also with the insight of AF447 accident .. you must think that if they avoided the CB plume by hundred miles .. the schedule was a joke .. but the passenggers and crew will be alive today .. commenting about their delay at arrival .....

[Clandestino]

How stall warning didn't supercede that condition ... ?That question should be and will be forwarded to human factors experts.

If you'd like to suggest that stall warning should automatically stop the trim dead, it's a bad idea. If THS helped one get near or into the stall, last thing he needs is frozen trim - it must help one out of the edge of envelope too. I simply can not mention name of David Pettit Davies - a superb pilot, brilliant writer and overall a great man - enough times in this thread.

So certainly with the insight of the AF447 accident you must think that the pilots were in very bad luck for have the 3 Pitot tubes frozen dead ..You have correctly interpreted part of my post. We know that "bad" ice that sticks to heated metal parts of the aeroplane, like TAT probes, pitots and engine compressors can be found in turbulent area near CBs but chances of finding it in aforedescribed areas are one to unknown, unknown being estimated as quite high. We only have anecdotal evidence that one type of probes is more often affected by "bad ice crystals" than other. We don't know why they get blocked but we're replacing them to avoid further incidents, which is quite reasonable and does not imply that probes are faulty. They just cannot cope with conditions met only very rarely in flight and there is no chance of replicating them in test conditions.


And also with the insight of AF447 accident .. you must think that if they avoided the CB plume by hundred miles .. the schedule was a joke .. but the passenggers and crew will be alive today .. commenting about their delay at arrival ..... That's a severe misinterpretation, sir. Flying is extremely hazardous enterprise and while appearance of thunderstorm clouds near desired route brings hazard level one notch up, overall risk increase is only slight. Sure, those CBs in ITCZ will certainly kill you if you fly into them but only those ignorant of way aviation works will make a drama out of it. There are hundreds of flights daily through ITCZ , there are thousands of deviations around CB every day and every now and then loss of airspeed indications ends up as brief note in airline safety bulletin, passengers being largely ignorant that their flight was anything but ordinary. So we as pilots do everything right most of the time.

Icing of the pitots was freak incident. Crew's reactions were unbelievable. Investigation moves into HF field.

There's excellent editorial in jul-aug issue of Aviation Safety World. Everyone, do yourselves a favour and read it.

[infrequentflyer789]

3Hole
Don't you see a problem with an aircraft with an automatic trim system that trims silently? For years we have had alerting devices on THS trim movement (clackers & whoolers) to let you know the thing is moving. Then comes the new guy on the block with a "better" idea. (Airbus)


And yet autotrim (up into a stall) is still catching out boeing pilots too. The bus (outside of direct/boeing law) will at least wind the trim back down if ask for nose down. Wheras it seems the 737 at least will wind the trim up to the stall warning and then leave the pilot to remember (or not...) to unwind it in recovery. Which is "better" ??


But no one saw the trim move. No one heard it move. If they had thought about it, they should have anticipated it moving, but they had other problems and it bit them.


The trim would have moved straight back down had they requested nose down beyond what the elevators could deliver. They didn't. They needed what 30deg(?) nose down to recover - never asked fro anything close. There was no stall recovery attempt that was stopped by the trim, there was no stall recovery attempted - period.

Had the trim movement kicked the crew in the rear or slapped them in the face for every degree it would have made no damn difference - they would have said yes trim me up, I'm trying to climb and the elevators aren't working... Had the THS ignored them and (say) gone to neutral, they would have overidden it with elevator all the way down. Had they had spotted it ignoring them, they would probably have manually trimmed back up (and what should the poor plane do then?). Why? Because they wanted to climb.

They were pitched up 15deg, falling out of the sky, plane not responding to controls [also low airspeed but they aren't sure on that]... and they had no clue they were stalled. Never mentioned. No (verbal) response to the stall warning - except perhaps "I’m in TOGA eh" [so I can't be stalled?] Throttles forward, stick hard back, why aren't we climbing ? "But we’ve got the engines what’s happening"

Sure, make the trim wheel clack, and in this case it's going to do all it's clacking during that 1min continuous stall warning. They didn't hear that, why would they hear the trim clack ?

[HarryMann]

Sure, make the trim wheel clack, and in this case it's going to do all it's clacking during that 1min continuous stall warning. They didn't hear that, why would they hear the trim clack ?

I think you lumping all the crew together... a few of us here (maybe a lot) are pretty sure that PNF was heads down while some of this was going on and not aware of long term NU inputs, or position of THS. Clacking may well have garnered his attention to NU trim and thus validity of SW... so yes, I anyway, disagree with you.

But as RRR suggested, a stick shaker and even subsequently a pusher may have been the only way with that crew on that night, in that aircraft

[airtren]

Clandestino,

I've read your recent posts with interest. I appreciate your sharing your experience and opinions.

Regarding the Stall, Stall Warning and THS, here are a number of clarifications:

As the operational recovery procedure from a Stall has been already revisited post AF 447 accident, there is a growing sentiment that it is time, for a closer look based on AF 447 data, and prior data as well, also to the automation being involved prior and during Stall, involving such elements as THS - the AF 447 case is bringing additional elements to those existing from previous incidents, documented by BEA reports.

BEA reports of incidents involving Airbus Stalls show that often the Stall Warning was active in parallel with the Autotrim moving the THS in max NU position, so the AF 447 case is not unique.

In some of the few Stalls that were followed by successful recovery, the BEA reports show that pilots struggled aggressively and valiantly with the combination Autotrimmed THS at max NU + Elevators max ND, which was a considerable obstacle to defeat.

On reflecting, it seems that in a Stall, a THS max NU, if anything, is only diminishing or delaying the effect of the Elevators ND, at a time, when the effectiveness of the Elevator ND is crucial and has stringent time constraints.

A successful Stall recovery involves a transition from Stall/NU, to ND with Elevators ND to regain speed, with loss of altitude, followed by Elevators NU, to level, and than regain altitude. It seems that in such conditions, having the THS stable in an Optimal Position from a Stall Recovery perspective - the Neutral position - which makes Elevator NU, or ND more deterministic, is a much better choice than having it move slowly from one position to another, under automation control, at times when the effects of Elevator ND, and/or NU have to be maximum and immediate, and are under very stringent time interval constraints.

It is more and more clear (to me) that the THS has a slow, longer time range stabilization function, which does not fit well with the short duration and requirements of quick transitions from NU/Stall, to ND and than NU during the Stall recovery, therefore, it is much better to have it neutral during these transitions.

Lastly, the suggestion that seems to have circulated during the past several pages regarding the THS at Stall is that it should be brought automatically to the Optimal Position for Stall recovery - Neutral Position - as opposed to being stopped in its current position, whatever that may be - which seems to be your understanding of the suggestion.


That question should be and will be forwarded to human factors experts.

If you'd like to suggest that stall warning should automatically stop the trim dead, it's a bad idea. If THS helped one get near or into the stall, last thing he needs is frozen trim - it must help one out of the edge of envelope too.

[Machinbird]

Hi Infrequent,
So you are planning to tape over any trim-in-motion bicycle bells that they stick in your Airbus cockpit because you don't want to have a clue?

The AF447 PF may have been completely clueless, but if either the PNF or Captain had tumbled to the trim position, do you really think they would have persisted in ignoring the possibility of stall? You have to give people a chance to get the information they need, even if their attention is distracted elsewhere. Stealth change in the position of a major control surface is downright hazardous.

Safetypee recently posted some interesting links. One of these is an excellent paper by Donald A. Norman titled, "THE PROBLEM OF AUTOMATION" http://www.jnd.org/dn.mss/Norman-overautomation.pdf
One of the paragraph titles sums up a key element of the problem: "THE PROBLEM IS NOT AUTOMATION, IT IS LACK OF FEEDBACK."

Do you remember the FBW pilot's favorite question? "Whats it doing now?"
That is the issue. The Airbus does not speak or gesture very effectively. How can flight crews maintain understanding of what their aircraft is up to? How do we keep the crew in the loop?

[Mad (Flt) Scientist]

I can see no reasonable means of engineering a system which would "reposition" the THS as a result of being in a detected "stall".

The whole point of being in a degraded mode with reduced or no envelope protection is that the system can no longer rely on one of its primary protection inputs - alpha.

if, despite that lack of reliability in alpha the system were, in a degraded mode, to have authority to move the most power single surface on the aircraft through close to full authority ( NU to neutral is a LONG way) it would be utterly impossible to show the system as being safe - in the event of an erroneous AOA input, you'd still be exposed to a large stab runaway, in effect.

With the AOA reliability suspect, the system can no longer act, automatically, on AOA. All it can do is draw the attention of the crew to the apparent high AOA situation - shake, shake - and hope they, with their human reasoning and airmanship, can resolve the situation which is outside the design parameters of the FBW system.

Unfortunately, in this case, they couldn't or didn't.

[HarryMann]

MadScientist

Surely alpha AoA is NOT suspect in Alt2 degraded mode. In fact AoA is (or should be) one of the easiest and most reliable of all transducers and can easily be kept independent.. that's why I find that a fully dynamically balanced vane was black-balled below 60 kts... It should never be black-balled... it should be extremely lightly spring loaded to a nominal 0 or 2,3 degrees +ve - for a low speed oscillating case, an RMS value could be used, and finally disregarded if mean and peak vales exceed a given ratio... for airspeed to be in any way involevd in its output IMHO is wrong.

There are also fairly easy ways to measure AoA less accurately as a fall-back value, electronically using hot-wire anemometry, as well as other methods

[Lyman]

So, bear with me. Say the Bus had gone to DIRECT LAW. There was a contingent not that long ago pushing that position.

TRIM is out of the picture, (Only manual, and they didn't notice it in AL)
A Stall surely, and a recovery with or without the PF's stick. He picked up the Roll pretty quickly, would he have grokked the situation in DIRECT? Would he have held Back ss through the 14 or so phugae? Or would he have recovered. Probably a dumb question. Except... in DIRECT, TRIM is optional. This STALL entry was unconventional... they never felt the joy of the nose dropping like a rock.

The Bus flies pretty well in the Stall......

[safetypee]

MFS, (agreeing with HM # 145), in this instance the AoA input might have been affected by a suspect ADIRU; there is no evidence of that, but there is valid supposition that the crew disbelieved the AoA derived stall warning.
However, the irony is that the requirements for an independent highly reliable stall system warning implies that AoA would be available for both stall warning and THS reposition – even an auto stall recovery (akin to a super stick push). The mechanism might be similar to existing Mach trim functions in the flight control computer.

[airtren]

MadScientist,

With the AF 447 BEA Report at hand:

Based on the BEA graphs, in the AF 447 case, the start of the Autotrim THS move towards max NU coincides (pretty much) with the start of the constant/continuous 53 seconds Stall Warning. During that period, from the beginning and close to its end - perhaps for about 40 seconds, the graphs also show NON-oscilatting (i.e. stable) AOAs emitted by IRSi(i=1,2,3) (page 111).

In this case, the system didn't really have to move the THS, it only had to stop the Autotrim from moving it from its -3 degrees position.

Back to the suggestion, perhaps there is a misunderstanding.

The suggestion in essence was not intended as to provide more automation during a Stall, but rather slightly change it, by reducing or eliminating the automation of the THS, in association with placing the THS in the "most optimal" position for the Stall Recovery. What the latter means, in terms of "optimal position" can certainly be the subject of more studying, but as a start can be considered THS Neutral.

Lastly, resolving the stability of the Stall Warning may provide the additional stability for the THS mechanism.

I can see no reasonable means of engineering a system which would "reposition" the THS as a result of being in a detected "stall".

The whole point of being in a degraded mode with reduced or no envelope protection is that the system can no longer rely on one of its primary protection inputs - alpha.

if, despite that lack of reliability in alpha the system were, in a degraded mode, to have authority to move the most power single surface on the aircraft through close to full authority ( NU to neutral is a LONG way) it would be utterly impossible to show the system as being safe - in the event of an erroneous AOA input, you'd still be exposed to a large stab runaway, in effect.

With the AOA reliability suspect, the system can no longer act, automatically, on AOA. All it can do is draw the attention of the crew to the apparent high AOA situation - shake, shake - and hope they, with their human reasoning and airmanship, can resolve the situation which is outside the design parameters of the FBW system.

Unfortunately, in this case, they couldn't or didn't.

[USMCProbe]

Lyman;
I am not sure how ACARS would report the FPV not working. I suppose it is possible. The Bird is not a "thing", it is calculated from the IRUs, and represents the course, both vertically and
horizontally, that the aircraft is moving. It is completely unrelated to any pitot static system, and does not rely on AOA measurements either. There are 3 IRU's on a 320 (maybe one more on a 330?). Any one of them could calculate the FPV.

The ACARS could report that the IRU's were not working, which would mean the FPV is not working, but I don't remember that from the data that I have read. It was only pitot-static information that was bad, for a short time, during the accident. At least that is what I have read thus far.

I flew 3 different jet aircraft with AOA as the primary performance instrument. The last one also had a HUD with the FPV. The great thing about the FPV on a PFD is that it gives you instantaneous information, including AOA, irrespective of what is going on with the pitot static instruments, and does not require any additional "scan" . It is where you are looking most of the time anyway.

[rudderrudderrat]

Hi USMCProbe,

It is where you are looking most of the time anyway.
Most of the time - it is turned off because the SOP is to use FDs. If you turn the FDs off, you then have to press another button to turn the PFV on.
If you are stalled, (with AoA in excess of 45 degs) it will be out of view (off scale below the visible part of the the PFD).

Apart from that - it's very simple to use.

[USMCProbe]

Exactly my point. If it is more than 8-10 degrees down from the pitch bars, your AOA is too high. Push forward until it is 2-7 degrees and keep it there. If it is "PFD-limited" i.e. off scale at the bottom, you are most definitely stalled.

And yes most of the time the "bird" is off , but they had over 3 minutes to hit the button and call it up. They were not trained to do this, and it would have probably saved the plane if they had been.

I would like to hear more about the ACARS report saying that the FPV was not working. Does anybody have any info on this? I can't believe a software generated symbol would be "monitored" and reported. The IRU's yes, almost for sure.

[AlphaZuluRomeo]

Hi

The suggestion in essence was not intended as to provide more automation, or a different automation during a Stall, but rather reduce or eliminate the automation of the THS, in association with placing the THS in the "most optimal" position for the Stall Recovery. What the latter means, in terms of "optimal position" can certainly be the subject of more studying, but as a start can be considered THS Neutral.

I agree with the first of your ideas (reduce or eliminate the automation of the THS) but not with the second part (repositionning the THS to neutral when stall warning is ON), as I feel unconfortable with this last idea.

Why not "simply" prevent any further NU movement of the THS by the auto-trim while the stall warning is ON ? This way :
- you prevent the automation to make the situation harder/longer to solve (THS full up = reduced ND effectivness = bad when stalled)
- but you don't inhibit/freeze totally the auto-trim (if you apply enough ND stick, the THS will eventually move ND too)
- you don't prevent the crew to manually apply (trim wheel) more NU or ND trim depending on their assessment of the situation
- you don't add another alarm (USE MAN PITCH TRIM type) to an already stressed crew to process.

Am I missing something? :)
I'm not sure that it will have changed anything substential in AF447's case, but I thought it was worth thinking about it.

On a side note, re the clues the crew had about the THS position : There was the trim wheel index, but also the F/CTL page on the center screen (showing all the control surfaces state, including numerical value for the THS angle). That page was automatically called due to some of the failures AF447 encountered (see ACARS analysis in interim reports #1 & #2).
It was not enough for the crew to notice. Sensory overload I guess?
I then support the idea of a clicking trim wheel, as it may give "another chance" to notice.

[AlphaZuluRomeo]

I would like to hear more about the ACARS report saying that the FPV was not working. Does anybody have any info on this? I can't believe a software generated symbol would be "monitored" and reported. The IRU's yes, almost for sure.

From BEA's interim report #2, ACARS analysis:
FLAG ON CAPT PFD FPV and FLAG ON F/O PFD FPV (2 h 11)
Symptoms:
Disappearance of the FPV (bird) on the PFDs, Captain and First Officer sides, and display of the corresponding flag.
Meaning:
This message indicates that the flight path vector (FPV) function is selected but unavailable. In order to lose completely this function, which is elaborated by the three IRs, in a way that is compatible with the CFR, one of the following three conditions must be met for each ADR:
¤ barometric vertical speed higher, as an absolute value, than 20,000 ft/min,
¤ true air speed higher than 599 kt,
¤ measured calibrated airspeed lower than 60 kt.
Once the operating conditions are satisfied again, the FPVs reappear on the PFD (if TRK/FPA mode is still selected).
@ 2h11, for the first time, as CAS lower than 60kt was measured.

[rudderrudderrat]

Hi AZR,

Thanks for the reference.

So despite being airborne (no weight on wheels, RAD Alts > 0 etc.) AB designed a system which removed both the stall warning and the FPV when IAS<60 kts.
What a wonderful system.

[JD-EE]

Trick is to have each and every airline pilot remember them and perform them flawlessly when his body clock says its 4AM, when he's jet-lagged, when he hasn't handflown the plane above FL100 for ages (if ever), when he does less than half a dozen manual landings per month, when he has never experienced control laws degradation and protection loss in real life and only seldom in simulator, when his knowledge of the principles of flight is lacking as he was only made to memorize the multiple choice answers to pass ATPL exams, etc... Whoever patents practical solution to this problem will quickly become millionaire.

Communications is the solution. (Actually, it solves an incredible lot of problems most people don't think of.)

When the auto-throttle and auto-pilot disconnect and Otto says meatware has the plane PF should start to execute the correct (UAS) drill AND PNF should start reading out the procedure to be checked off one by one. The PNF should call it out loud and clear. PF should repeat loud and clear.

If they face a memory drill, PF may act, but PNF must call out the memory procedure and PF must acknowledge with the full step's instructions. Then you have two human brains working on the same problem in case one flies South for the winter or something. If PF fails to acknowledge PNF swats him over the head with two day old pizzas. If PNF fails to call out the drill, PF shouts obscenities about his 33rd cousin's maiden aunt - anything to get the communications cycle running with check and cross check.

It's so darned simple it's not even patentable.

I tell you three times, Communicate, Communicate, Communicate. Therefore it must be true. (Sorry Mr. Carroll for mangling it. But, as a dirty job it had to be done.)

[rudderrudderrat]

Hi JD-EE,

Communications is the solution.
I agree when conducted in a calm, relaxed atmosphere. Verbal communications on top of the plethora of cockpit noises, causes some noises to be filtered out. (Stall Stall??).

The big advantage of connected control wheels, is Pilot A Knows what Pilot B is attempting to do without a word being spoken.
AB have designed a flight deck where the crew are separated by their control sticks. They may as well be in separate rooms talking to each other on the telephone whilst watching their individual Flight Displays.

[JD-EE]

Lone, if the plane has been showing a ground speed of say 475 kts and an airspeed of 275 kts. The airspeed goes away because somebody insulted the pitot probes. If the ground speed drops but stays above 200 kts should the plane decide it doesn't know AIR speed so it should not impolitely sneer at the pilot about the plane stalling? Seriously, simple loss of airspeed is not a total loss of the ability to make a usable guess of airspeed for at least a few minutes with regards to stall warnings in a sane world.

Unfortunately, in the case at hand, I really don't think that would have made a difference. PF lost his head through lacunae in his training, unrecognized fatigue, or some other factor we can only guess about. PNF was not being communicative enough calling out proper procedure so that the cockpit TEAM were working on the same problem with their full brains. (And don't forget that two day old pizza for waking up the PF.)

[JD-EE]

A33Zab quoted: The altitude indications are based on GPS data. Two amber dashes cover the last two digits because the GPS altitude is less accurate than the barometric altitude.

I can (sort of) understand that back when the Denial Of Accuracy or dithering or whatever you want to call it was enabled by default on GPS. 100 meters was about as good as it got. Now it's about 10 times better between the DOA going away (I had a paw or two in that design) and basic improvements in clock and ephemeris accuracy. I wonder if TPTB have recognized this and given GPS a slightly better reputation and treatment.

Of course, I am presuming they have the intertial platform providing aiding data to the GPS receiver to help it track through extreme turbulence and other high dynamics conditions.

[JD-EE]

airtren and others, if meatware pilots are so darned good, why do you and the others believe the FBW aircraft are consistently showing very significantly lower accident rates when it has just been demonstrated that current levels of flight training have deteriorated significantly? Could it be that the automation actually provides a serious added benefit?

[CONF iture]

And yet autotrim (up into a stall) is still catching out boeing pilots too. The bus (outside of direct/boeing law) will at least wind the trim back down if ask for nose down. Wheras it seems the 737 at least will wind the trim up to the stall warning and then leave the pilot to remember (or not...) to unwind it in recovery. Which is "better" ??
The one that keeps things simple, always :
In manual flight, don't expect anything to trim for you - You, the pilot, are responsible for it.

Had the trim movement kicked the crew in the rear or slapped them in the face for every degree it would have made no damn difference - they would have said yes trim me up, I'm trying to climb and the elevators aren't working... Had the THS ignored them and (say) gone to neutral, they would have overidden it with elevator all the way down. Had they had spotted it ignoring them, they would probably have manually trimmed back up (and what should the poor plane do then?). Why? Because they wanted to climb.
Leave to the pilot the responsability for this type of action ... and if he does actually choose to trim up ... then I won't show up to defend his action.

If you'd like to suggest that stall warning should automatically stop the trim dead, it's a bad idea.
No please, not even more automation.
At the earliest sign of disagreement between the probe values, Airbus should say :
As I'm not sure exactly what's happening, I auto cancel all my wonderful features, and I give you back a conventional aircraft in direct law.

[JD-EE]

Remove from the plane the Pitot tube and you have no more speed measuring system at all

Brown soft stinky material such as emanates from the South end of a North facing fertile male bovine.

The inertial and the GPS systems are still working. They may not know your air speed within 60 kts and your altitude likely well within 100'. BUT, it does know the plane had better not be touching the ground when the ground speed is over say 200 kts or even only 120kts and if the AoA is excessive for the given ground speed politely slap the pilot with a dead fish and tell him he's stalled. (Which would have done no good because the PF thought he was doing the correct stall drill, apparently. Where he dreamed it up, nobody quite knows.)

Furthermore, if the last pitot reading was 275, the pitch of the aircraft is about normal, and the thrust is about normal you can pretty decently presume 5 minutes later you're still going roughly the same speed at roughly the same altitude even if the GPS set has gone toes up.

Now, the design of the plane may have hidden some of this data from the PF or the PF had not been trained to find and use it. But the plane bloody well knows in one of its computin' thangs.

[JD-EE]

infrequentflyer789 (http://www.pprune.org/members/214043-infrequentflyer789), you and others may be on to a distinct problem with the THS concept. A clacker might be a good idea.

But, in the case at hand I don't think it contributed anything to the problem other than making bad a little worse. Track the elevator position from the peak of the climb to ocean surface. Did it ever, even once, go nose down rather than slightly less nose up? Had the PF gotten the elevator past neutral into ND territory the THS would have started to unwind, fairly smartly. And we'd be arguing if it had unwound as fast as it should have or whether it should have wound itself up to the stops. But the problem visible is that PF had nose up TOGA on his mind and nothing else penetrated.

That is why communications and team work are needed in the cockpit. The team members must communicate with each other what they are doing and what they are expected to be doing. (And 2 day old pizzas must be kept in stock for waking pilots out of their trances.)

[RetiredF4]

airtren and others, if meatware pilots are so darned good, why do you and the others believe the FBW aircraft are consistently showing very significantly lower accident rates when it has just been demonstrated that current levels of flight training have deteriorated significantly? Could it be that the automation actually provides a serious added benefit?

You are correct, and nobody doubts the benefit of automation in itself.
But it can and must further improve and subjects asociated with automation must be looked at with more attention than before..

Or are you saying, automation is at the end of its developement, that is as far as we can get, human machine interface is perfect and that will be for the next 100 years?

[airtren]

AlphaZuluRomeo, thanks for your reply, and ideas.

Hi

I agree with the first of your ideas (reduce or eliminate the automation of the THS) but not with the second part (repositionning the THS to neutral when stall warning is ON), as I feel unconfortable with this last idea.

Why not "simply" prevent any further NU movement of the THS by the auto-trim while the stall warning is ON ?

The reasoning for returning the THS to a Stall Recovery Optimal Position - I picked Neutral as that seems the best candidate - is that from other BEA reports it seems that in other Stall incidents, unlike AF 447, where the THS was started the move NU after the Stall, the THS was further NU, or max NU already before the Stall.

A THS Stall Recovery Optimal Position is one, which yields the most effective Elevators ND, and NU actions, which are needed in a (quite) quick sequence during Stall Recovery .

- but you don't inhibit/freeze totally the auto-trim (if you apply enough ND stick, the THS will eventually move ND too)
As long as the Stall is ON, as it is a slow move/reacting surface, I would not risk interference with the Elevators, which need be quickly very effective, and for short time intervals.


- you don't prevent the crew to manually apply (trim wheel) more NU or ND trim depending on their assessment of the situation
- you don't add another alarm (USE MAN PITCH TRIM type) to an already stressed crew to process.
Manual control of course remains. Sound alarm can become easily a disturbance, so agree. A visual - a LED - would, IMO, be useful.


I'm not sure that it will have changed anything substential in AF447's case, but I thought it was worth thinking about it.
The Elevator NU would have been less effective, while the Elevator ND would have been more effective. Which means different pitch angles and AoAs - less NU, more ND. That we know. This could be simulated. Would they have taken advantage of it?

[GerardC]

JD-EE:
Communications is the solution. (Actually, it solves an incredible lot of problems most people don't think of.)

When the auto-throttle and auto-pilot disconnect and Otto says meatware has the plane PF should start to execute the correct (UAS) drill AND PNF should start reading out the procedure to be checked off one by one. The PNF should call it out loud and clear. PF should repeat loud and clear.JD-EE, you are spot on.
Communication IS the solution : communication between pilots AND communication between the plane and the crew.

What do we have (see BEA's #3 report page 29) :
- at 02:10:05 speeds are OK and Otto decides to quit (why ?) ;

- at 02:10:06 speeds are OK ; plane starts banking ; PF : "I have control" ;
Wouldn't a 'bus pilot suspect some sort of FBW computer(s) malfunction and start "aviating" ?

- displayed speeds start to go wrong at 02:10:07 : after the, so far unexplained, FBW malfunction. (Cause or consequence of the FBW malfunction ?)

Now look at pages 45 and 46 :
- page 45, BEA tries to make you think that the "NAV ADR DISAGREE" message came immediately after the "AUTO FLT AP OFF" message. This is not true. "NAV ADR DISAGREE" came in only at 02:12:XX (why XX ?)

- page 46, order of messages :
1) "AUTO FLT AP OFF" ;
2) "NAV ADR DISAGREE" suggesting "IF SPD DISAGREE -> ADR CHECK PROC... APPLY" ;
is only shown to the pilots at 02:12:XX when situation is already largely out of control.

Questions :
- why should the PF call for unreliable speed drill at 02:10:05 when Otto quits and speeds are OK ?
- in an event leading to AP disconnection, why is such a smart AP/FD system unable to maintain for a few seconds the average pitch of, say, the past 5 or 10", until the proper ECAM alert message is displayed ?
- am I the only one to think that the "NAV ADR DISAGREE" ; "IF SPD DISAGREE -> ADR CHECK PROC... APPLY" message displayed at 02:10:08 THEN Otto disconnection at 02:10:10 plus a stick shaker instead of an aural stall warning could have saved 228 lives that night ?

Better communications and smarter design are solutions for a better flight safety.

[AlphaZuluRomeo]

@ rudderrudderrat : My pleasure :)

Regarding your comment:
AB designed a system which removed both the stall warning and the FPV when IAS<60 kts.
What a wonderful system.
The thing is AoA probes (needed for stall warning & FPV) are (or deemed) unreliable when IAS<60 kts. I'm not sure an unreliable FPV and/or stall warning is better than none at all?

About the PFV: if I read correctly the annex 4 (parameters) of the 3rd interim report, HDG-V/S was selected on AF447, never was TRK-FPA asked by the crew. Therefore, the unavailability of the bird, although damageable in absolute terms, would have changed nothing in the case of AF447.

[AlphaZuluRomeo]

Hi airtren

The reasoning for returning the THS to a Stall Recovery Optimal Position - I picked Neutral as that seems the best candidate - is that from other BEA reports is that in other Stall incidents, unlike AF 447, where the THS was started the move UP after the Stall, the THS was further UP, or full UP already before the Stall.
I understand your point, it is valid but I still feel uncomfortable with it, because:
- it adds another layer of complexity to be aware of (KISS)
- it may be dangerous if stall warning isn't 100% reliable (and it's not)
- the crew must react to the stall by pushing ND. That would give ND elevators, then ND THS if the latter was NU by far.
(note : on AF447, if one accepts that the flight control surfaces acted accordingly with the crew inputs, the crew never pushed enough on their sidestick to make the elevators go ND, only less NU)

As long as the Stall is ON, as it is a slow move/reacting surface, I would not risk interference with the Elevators, which need be quickly very effective, and for short time intervals.
I don't think I understand which interference you may be refering to? I didn't suggest to limit the elevators in any way, only to prevent THS NU move.
I agree that if THS stays where it is, the recovery may be a bit longer than with the THS neutral (or ND), but as the crew has to push SS ND, elevators will go ND and eventually THS too.

[airtren]

jd_ee,

Benefits of technology progress are so pervasive that IMO go without saying. Your reminder is welcome. It is also an opportunity of another reminder:

Somebody already said this on these threads a lot better than me: Technology progress changes the safety parameters practically in every field. Technology progress is manifesting itself not only in new systems, but in improvements of existing ones as well.

Would I be reading too far in your post, that your thinking is that existing automated systems don't need improvements? for an improved safety?

airtren and others, if meatware pilots are so darned good, why do you and the others believe the FBW aircraft are consistently showing very significantly lower accident rates when it has just been demonstrated that current levels of flight training have deteriorated significantly? Could it be that the automation actually provides a serious added benefit?

[Linktrained]

Who reduced the Thrust from TOGA to Flight Idle for a few seconds near the top of the climb, when the pitch of the aircraft reduced, slightly ?
Then TOGA was restored, for the all that remained of the flight.

PF held/ had held a French Gliding Licence from a few years earlier. I do not know how active he had been or still was. A Brevet "B" only required a couple of "S Turns" and a safe landing. I imagine ( but I do not know) that our PF went well past this level, to Brevet "D" or beyond...

(AF encouraged some former members of Cabin Crew to become Pilots - our Captain had been one, earlier in his career.)

The PF could have become a gliding Instructor. He probably would have had aero-tows with the cable fixed to the nose of the glider. This would have given him enough height to practice stalls and spinning. He would have been winch launched, sometimes, where the cable might be fixed lower down, perhaps under the Pilot's seat, to get the best height possible, rather like a kite.

Underslung engines have an effect on the pitch of an aircraft, normally advantageously " more power=nose up..... less power= nose down"

At this moment, for AF447, the nose ought to come down.

Someone else said that only Nostradamus would really know what was in PF's mind. But who did alter the Thrust, twice?

[AlphaZuluRomeo]

Hello GerardC
- at 02:10:05 speeds are OK and Otto decides to quit (why ?) ;
Did you read §1.16.4.1 of the 3rd interim report?
Also remind that:
- recorded speed are from pitots/ADR 1 & 3
- AP2 was ON, it uses IIRC pitot/ADR 2 as primary

- at 02:10:06 speeds are OK ; plane starts banking ; PF : "I have control" ;
Wouldn't a 'bus pilot suspect some sort of FBW computer(s) malfunction and start "aviating" ?
I don't quite understant your question. :confused:

- displayed speeds start to go wrong at 02:10:07 : after the, so far unexplained, FBW malfunction. (Cause or consequence of the FBW malfunction ?)
See above ; I agree with you if you replace "displayed speeds" (there was 3 of them) by "recorded speeds" (only 2).

Now look at pages 45 and 46 :
- page 45, BEA tries to make you think that the "NAV ADR DISAGREE" message came immediately after the "AUTO FLT AP OFF" message. This is not true. "NAV ADR DISAGREE" came in only at 02:12:XX (why XX ?)
BEA tries what? It's written plain & simple:
The table below lists the messages generated during the event, in order of priority and in a form similar to that in which it was displayed to the crew.
The purpose of this table is to show which message(s) will be directly seen by the crew, as the screen cannot contain them all.
That's why on the next page, with the hypothesis that no message was ever acknowledged/erased, the BEA shows simulated printscreens on a chronological basis (other purpose).

[OK465]

From various posts:


If it is "PFD-limited" i.e. off scale at the bottom, you are most definitely stalled...

And yes most of the time the "bird" is off , but they had over 3 minutes to hit the button and call it up.



Most of the time - it is turned off because the SOP is to use FDs. If you turn the FDs off, you then have to press another button to turn the PFV on.


...HDG-V/S was selected on AF447, never was TRK-FPA asked by the crew.

FPA/TRK was selected at some point or they would not have had FPV FLAGS reported on the PFD.

FPV was flagged because all three ADR's were simultaneously unreliable at various times. With Airbus you need at least one good ADR to get an FPV. It is not totally inertial dependent. At any time at least one ADR became reliable, the FPV would have returned (possibly not visible), if selected. On a B737 you can fail all (both of 2) ADR's and still have a full-time FPV. It is totally inertial.

The FD does not have be turned off to select FPA/TRK. When FPA/TRK is selected, the HDG/VS dual cue steering is replaced by the FPV FD symbol and the "bird" (FPV) is "displayed".

If the FPV is offscale of the PFD when it returns, how do you know what its status is?

It is not "ghosted" or caged at the lower limit (bottom) of the PFD like a HUD FPV. It is GONE.

[AlphaZuluRomeo]

FPA/TRK was selected at some point or they would not have had FPV FLAGS reported on the PFD.
My mistake, then. What you write is consistent with the text of the 3rd report (§ 1.16.6.3, just re-read it), but there is an inconsistency with the annex 4 (parameters) on p. 108 (bottom) which shows only HDG-V/S selected. Perhaps the sampling period isn't tight enough for this parameter? Or the "for EIS2" mention means something I don't get at the moment (any help appreciated)?

[A33Zab]

This 3rd report is an interim report like the 1st and 2nd report.
You need to read al reports!, some questions you have are already answered in the previous reports.


after the, so far unexplained, FBW malfunction


Nothing to explain here, there wasn't a FBW malfunction.


and smarter design are solutions for a better flight safety


but what if a smart design is not understood and not properly used by the user? like AF477.
One can say there is a design flaw or not smart enough systems during the design phase or after the first years of operation (like DC10 aft cargodoor).
But not after so many years of operation, by then there are other issues,
and that may be clear to all of us by now.

Pilot error due lack of training, absence of CRM and TEM.

Isn't there room for technical improvements?

Yes off course there is, but not tactile feedback (which is of no use in FBW) or introduce other protections (THS inhibit at high AoA) because there you ask for other problems.

A. already made changes to AoA BEFORE this event and offered this as option. 'BUSS'
Svoice 'STALL STALL' would have sounded until impact.

Would this have prevented this accident?
proberly not because it wasn't understood in the 54s it was present!

AF (like most operators) didn't opt for that mod because there wasn't a reason to do so before.
Now things have changed and this 'BUSS' option may become standard (like it is on A380 and will be on A350).

[airtren]

AlphaZuluRomeo, thanks for the follow up.

The Stall Warning reliability is among the concerns raised with the AF 447, here on this Forum, as well as the BEA report. I am confident, or perhaps just hoping that it is going to result in a BEA recommendation, and will be mandatory to address.

I understand your perspective, and I perceive it, as being in the same direction.

Perhaps with the risk of repeating myself, I see in general the time constraints of the Stall Recovery as a very important parameter. There is very little time for such a Stall Recovery - we know it took about 3 minutes to loose 38K ft of altitude - and with the THS being a slower moving control surface, I see it beneficial to have the THS down to Neutral sooner, rather than latter, and let the Stall Recovery be handled with the Elevators as a main element.

Hi airtren
I understand your point, it is valid but I still feel uncomfortable with it, because:
- it adds another layer of complexity to be aware of (KISS)
- it may be dangerous if stall warning isn't 100% reliable (and it's not)
- the crew must react to the stall by pushing ND. That would give ND elevators, then ND THS if the latter was NU by far.
(note : on AF447, if one accepts that the flight control surfaces acted accordingly with the crew inputs, the crew never pushed enough on their sidestick to make the elevators go ND, only less NU)

I am sorry for not being clear, or not using a better word. Let me try to clarify - sorry if there is any repetition.

The use of "interference with the Elevators" was meant as "diminishing the effectiveness of the Elevators".

Specifically, it was in reference to the effect of a THS NU, versus a THS Neutral when summed with the Elevators ND, for the first phase of the Stall Recovery (ND command, gain speed, loose altitude),

efficiency THS NU + ElevatorsND < efficiency THS Neutral + Elevators ND

respectively the effect of THS ND, versus a THS Neutral when summed with the Elevators NU, for the second phase of the Stall Recovery (NU command, level off, stabilize and start regaining altitude).

efficiency THS ND + ElevatorsNU < efficiency THS Neutral + Elevators NU

In both cases, the THS Neutral is more "neutral" :) than the others.

An important consideration was that while it is true that the non-neutral effect of the THS may mean "amplifying", as opposed to "diminishing" - for instance THS ND with Elevators ND, and THS NU with Elevators NU - for the duration of one of the two phase of the Stall Recvery, it is also true that the Slow speed of the THS move, is a factor that may transform the "amplifying" during phase 1, into a "diminishing" in phase 2, at the transition from phase 1 or phase 2 of the Stall Recovery.

The short time available for the Stall Recovery, requires a very quick. and very effective ND to NU transition from phase 1 to phase 2 (described above), with fully effective Elevators - any "diminishing" is undesired.

I hope it is clearer.:)


I don't think I understand which interference you may be refering to? I didn't suggest to limit the elevators in any way, only to prevent THS NU move.
I agree that if THS stays where it is, the recovery may be a bit longer than with the THS neutral (or ND), but as the crew has to push SS ND, elevators will go ND and eventually THS too.

[Clandestino]

a few of us here (maybe a lot) are pretty sure that PNF was heads down while some of this was going on and not aware of long term NU inputs, or position of THS.How low must one's head be to not notice the attitude and altitude on his PFD? Not to hear "STALL STALL STALL"?

But as RRR suggested, a stick shaker and even subsequently a pusher may have been the only way with that crew on that night, in that aircraft Stick pushers are not there to deal with crew unaware of what goes on around them. They are potentially very dangerous devices and are fitted only to aeroplane that are considered to be more dangerous without then with them. DP Davies has written an excellent treatise on stick-pushers, pages 130-140. His analysis of risks and benefits connected with pushers is amazingly applicable to any other envelope protection too.

As I'm not sure exactly what's happening, I auto cancel all my wonderful features, and I give you back a conventional aircraft in direct law. And give you direct stick-to-elevator? Half up elevator at mach 0.8 would give pretty much the same outcome we have now though the path to it might be somewhat different and perhaps include a lomcevak or two.

Communications is the solution. There was nothing to communicate except utter bafflement in the cockpit. Crew got way behind the aeroplane, completely lost SA and never regained it during final minutes.

Could it be that the automation actually provides a serious added benefit? Absolutely. There are two dominant views on implications of "Swiss cheese" accident model. Optimistic theory claims each and every cheese slice has to be as hole-free as possible to reduce the possibility of accident to minimum. Modern management theory is that as any cheese layer stops the accident as well as any other, it is important to first make cheaper slices thicker and less porous. Others can be thinner, with more holes if number of the dead when all are breached falls below the statistically acceptable.

Automated aeroplane with competent pilots is safest of them all but do the additional costs of proper pilot training justify the increase in safety? If you really believe that life is priceless, they do but that's not the way thing work in our day and age. Yours, mine and everyone else's life have a price tag on them.

[CONF iture]

Half up elevator at mach 0.8 would give pretty much the same outcome we have now though the path to it might be somewhat different and perhaps include a lomcevak or two.
Non sense - Why would you pull the sidestick half way ?
Don't you adapt your inputs depending how your 'direct law' aircraft respond ?

[Clandestino]

Non sense - Why would you pull the sidestick half way ?Why would you pull half-way in ALTN2 when encountering UAS? The one who could answer that is unable to talk to us and I strongly suspect he would be unable to coherently explain what was he thinking and doing even if AF447 were somehow timely recovered from stall.

Don't you adapt your inputs depending how your 'direct law' aircraft respond ? Don't you adapt your inputs to achieve desired attitude, no matter what control law are you in?

[Lyman]

So again, a question. In AL the RTL protects the control, the Rudder. From your "Half elevator up at Mach.8." I assume you propose an inadvertent elevator input (or commanded, either way), might cause serious problems (lomcevak).

ROLL is DIRECT, RUDDER is DIRECT, but PITCH is NOT DIRECT, and it is UNPROTECTED.

The logic here escapes me, could you explain to me? I can understand the need for deflection protections, but why not in the elevators?

On the one hand, the airframe defends the controls, on the other , the Flight Path (elevators). Why?

So here: In a degradation to Alternate LAW, the a/c anticipates (with its logic) control inputs that may be excessive. Obviously, the need is to MANEUVER, NOT CRUISE. Why in Heaven's name is TRIM allowed at all.

It is TRIM, and not a primary FLIGHT CONTROL. keep it firmly at 0 degrees...... Throughout.

If the TRIM is active, it absolutely compromises the elevators' authority.

If ACTIVE, it is also sluggish (behind), and the Pilot must form a different response to variable combinations of TRIM/ELEVATOR, which he cannot see, feel, or "hear". This is not a recipe for disaster? One that can be blamed on "Lack of Handflying?" POOR TRAINING? ETC?

[rudderrudderrat]

Hi Clandestino,

Don't you adapt your inputs to achieve desired attitude, no matter what control law are you in?
Yes you do. But suppose you are "overloaded" by warnings as you struggled to maintain wings level. Consider these two alternatives.
1) In Direct Law, the stab trim would not have moved. If PF relaxed the back pressure, the aircraft would have nosed down naturally whilst it attempted to regain it's trimmed speed.
2) In ALT LAW, when he relaxed back pressure, the aircraft maintained it's nose up attitude as the speed washed off.

In which Law do you think is easier to actually stall?

@AlphaZuluRomeo,
The thing is AoA probes (needed for stall warning & FPV) are (or deemed) unreliable when IAS<60 kts. I'm not sure an unreliable FPV and/or stall warning is better than none at all?
The IAS was grossly under reading. The acute angle of attack to the pitot tubes rendered them useless. The fact that the aircraft is airborne should be sufficient logic to maintain stall warnings and Angle of Attack.

[Clandestino]

If the TRIM is active, it absolutely compromises the elevators' authority.Well, duh. It is supposed to be that way. I don't mind writing for a zilionth time that DP Davies has it all neatly explained.

In ALT LAW, when he relaxed back pressure, the aircraft maintained it's nose up attitude as the speed washed off.Just when did he relax the backpressure? You are acquainted with sidestick traces from 3rd intermediate report, aren't you? A couple of times stick goes forward but quickly gets behind neutral again and stays there before elevators even have the chance to get to neutral.

[airtren]

A33Zab,


.
One can say there is a design flaw or not smart enough systems during the design phase or after the first years of operation (like DC10 aft cargodoor).

But not after so many years of operation, by then there are other issues,
and that may be clear to all of us by now.

The better understanding of possible problems later in the life of a complex system is somewhat contradictory with discounting the possibility of uncovering flaws later in the life of such a complex system. The latter calls for some reminders:

- the probability of uncovering all flaws/shortcomings/bugs during tests, or in the first year or years of operation of a system, is decreasing with the complexity of the system.

- a shortcoming may stay hidden for a good amount of time, as long as its hosting path in the system's algorithms is not operated, or fully operated - the type of AF 447 Stall is not happening often.


Isn't there room for technical improvements?

Yes off course there is, but not tactile feedback (which is of no use in FBW) or introduce other protections (THS inhibit at high AoA) because there you ask for other problems.Can you elaborate on "THS inhibit at high AOA", in terms of description, and the problems incurred? I am not sure if you refer to the elements that were the object of several earlier posts on this thread, but either way, elaborating would be helpful, and would be appreciated.

A. already made changes to AoA BEFORE this event and offered this as option. 'BUSS'
Svoice 'STALL STALL' would have sounded until impact.Are there any reasons for the BUSS not to have been adopted? particularly for a route that has a higher risk? like the AF 447's?

[rudderrudderrat]

Hi Clandestino,

You are acquainted with sidestick traces from 3rd intermediate report, aren't you?
Yes thanks.

See Page 74.
"The PF copilot said “I have the controls” and made rapid and high amplitude lateral, almost stop to stop, inputs. He also made a nose-up input that increased the airplane’s pitch attitude
up to 11° in ten seconds……."

The nose up request on the side stick rotated the aircraft by about 1 deg per second. If they had been in Direct Law, as the speed washed off, the nose would have wanted to drop and the back stick pressure would have been less effective. The aircraft may have stabilised at a speed above the stall whilst still holding the same back stick pressure.

Do you think it is easier to stall in Direct or Alternate LAW?

[Clandestino]

Yes thanks.Great. Now we'll deepen our acquaintance with sidestick and controls positions on page 108 which are raw data behind words on page 74 that you cling to. You have certainly noted that despite right sidestick hovering around halfway nose-up input, elevators that are delivering that one degree per second pitch-up are barely off neutral. What do you think would happen if FBW went into direct law and tried to move elevators to half-deflection at 0.8 Mach? I don't know whether it would be even faster zoom climb to stall at higher pitch, dynamic stall or overstress but it would certainly be uglier than 11° in 10 seconds pitch-up.

Do you think it is easier to stall in Direct or Alternate LAW? I have seen no valid argument on which to base my decision on the matter, therefore I'm undecided for the time being.

[rudderrudderrat]

Hi Clandestino,
I don't know whether it would be even faster zoom climb to stall at higher pitch, dynamic stall or overstress but it would certainly be uglier than 11° in 10 seconds pitch-up.

I don't know either - but I bet they would have felt it.
Instead of an unnoticed, increasing pitch up - all he would have had to do was let go and the aircraft would have naturally nosed down to recover any lost speed. It would still have been trimmed to fly at M.08.

[airtren]

Clandestino, Ruderrat,

An interpretation that the pilot has intentionally, in full knowledge of the situation entered NU commands, calls for a reminder of the confusion shown by the CVR transcript in the BEA report.

Furthermore, there is additional data to consider - such interpretations were advanced before on the AF 447 threads, but do not recall them being documented this way.

The combining of the Pitch and Roll control BEA graphs - see bellow - show a significant coincidence and persistence of Pitch and Roll commands.

As I think it is normal that not every Pitch command required a Roll, and vice-versa, this can be interpreted as:

a). a lesser familiarity with the separation of the two types of commands on the stick, and furthermore:

b) a lesser familiarity with the Neutral position of the stick.

Consequently, it seems that what would have been a perceived Neutral position of the stick for Pitch and for Roll commands, was in reality a OFF Neutral position, in reality a NU command, and respectively a Roll command.

In conclusion, this can be interpreted that the persistent NU commands were not intentional, but just a simple consequence of the handling of the stick.

If the NU commands were not intentional, than the Autotrim of the THS to max NU was not intentional.

Furthermore, the THS was positioned to the max NU position during the Stall Warning time interval, contrary to what the computer logic would have asked for, if the two logical paths (Stall, and Autotrim) were linked.

Note: if we expect the pilot to conclude that in STALL a NU commands is a NO, NO, why shouldn't we expect the computer to arrive at the same conclusion???

It is possible that the BEA work on the Human Factors and a future report will add more analytical elements, interpretations, conclusions, and recommendations that may intersect with elements presented in this post


http://farm7.static.flickr.com/6207/6059502869_187d8a4571_b.jpg

Well, duh. It is supposed to be that way. I don't mind writing for a zilionth time that DP Davies has it all neatly explained.

Just when did he relax the backpressure? You are acquainted with sidestick traces from 3rd intermediate report, aren't you? A couple of times stick goes forward but quickly gets behind neutral again and stays there before elevators even have the chance to get to neutral.

[Lyman]

Clandestino

It is NOT 'intended' to compromise the elevator's authority. Its intent, as you well know, is arguably out of place in Alternate LAW 2, when Pitch is unprotected, and the Pilot hasn't a clear knowledge of the actions of the THS.

Further, in a recovery, what the THS does more than anything, is FRUSTRATE the Elevator's authority, and you also know that.

To Stall in DIRECT? or ALTERNATE? Like clandestino, a toss up.

This I do know. To recover from a STALL? given 447's data, DIRECT. WHY? Because without a fully up THS, the elevators could not have kept her in the STALL (caveat: pending a final decision by BEA re: CG). She'd have dropped her nose, something these gents were not granted as a cue that STALL was underway, (and just broken).

If, after the nose dropped, the PF kept holding full back, a slow repetition of the same thing. Then if he eventually did not figure things out, perhaps the rollercoaster may have prompted one or the other worthies to sit down.

[airtren]

Clandestino,

I read your post as saying that "an active TRIM is supposed to be absolutely compromising the elevators authority".

Can you explain that in your own words, interpretation and logic?

Well, duh. It is supposed to be that way. I don't mind writing for a zilionth time that DP Davies has it all neatly explained.

If the TRIM is active, it absolutely compromises the elevators' authority.

[mm43]

The question should really be, "Did the elevator have ND authority during the period when the aircraft was on the wrong side of the stall warning?" The answer is yes it did, and indications are that it would have had control up to around 180 KCAS. If a Side Stick max ND command had been made and held for about 2 seconds, the THS would have started neutralizing the elevator demand. Any further ND would have provided increased control as THS followed and airspeed ramped up. The whole process could be sped up by use of the Manual Trim Wheel.

The only real problem I see in the way the THS operates is, the use of the Manual Trim Wheel appears to have been absent from the training regime, or at least discouraged. Why has this happened? The THS is a major control surface, and where it is positioned should be part of the instrument scan. OK, in Normal Law it is taken care of automatically, but in Alt Law the situation changes. Did the PF ever understand how the THS operated in Alt Law? We will never know, but if you don't understand a STALL WARNING, nor understand that consistently pulling the nose toward the heavens is not going to help you on this planet, then what chance does a trim wheel turning constantly for a minute within your periphial vision have of getting noticed.

As for the THS doing its 3°NU to 13.6°NU thing; it had nothing to do with 'autotrim'. It was operating in an integrated demand/time manner to neutralize the elevator input demand - nothing more. The elevator demand was wilful.

The A310 TAROM approach zoom climb/stall/recovery (http://www.bea.aero/docspa/1994/yr-a940924a/pdf/yr-a940924a.pdf) incident at Orly on 24 September 1994 provides some insight into the movements of the THS. Though I note the PNF made some manual adjustment to the THS following the stall recovery while setting up for their second approach, and without the PF being advised. That doesn't mean that the PF was unaware that the THS was moved to provide better trim. Also, the incident was marked by NO stall warning, and the reason given was that AoA SW thresh-hold had not been reached before the IAS had dropped below 60 KTS and the control logic inhibited it. Sound familiar?

Essentially the same logic with minor changes has been around for a long time. Unfortunately it requires incidents/accidents before everyone realizes how some of this control logic works.

[Clandestino]

Can you explain that in your own words, interpretation and logic?Yes.

Trimmable horizontal stabilizer, on any such equipped aeroplane, is always the most powerful control surface and easily overpowers the elevator.

Get this book:

http://covers.openlibrary.org/w/id/701216-M.jpg

and read from page 35 onwards for details.

[RetiredF4]

OK, in Normal Law it is taken care of automatically, but in Alt Law the situation changes. Did the PF ever understand how the THS operated in Alt Law?

What does change in Alt Law?
According to my references and also posted here from others in Alt Law the autotrim is working as well.
What are you getting at with "the situation changes"?

[jcjeant]

Hi,

What does change in Alt Law?
According to my references and also posted here from others in Alt Law the autotrim is working as well.
What are you getting at with "the situation changes"?Very little change ..
No more pitch protection
Autotrim is working as well .. but the change is that the elevators are now commanded by the pilot SS .. no the autopilot
As we had seen in the AF447 case .. this can make a difference

[Lyman]

No, jcjeant, the ss' inputs go to the computer, which controls the elevators. You are describing PITCH DIRECT, not part of AL2. PITCH DIRECT would make possible a lomcevak, perhaps, as the elevators are not "protected".

[airtren]

Hello mm43,

I've noticed in your post a number of differences in the use of words, that are perhaps just semantics? There are also a number of adjustments to be called. For instance:

... Any further ND would have provided increased control as THS followed .... The whole process could be sped up by use of the Manual Trim Wheel.

Your "THS followed" .... is the THS being moved NOT MANUALLY, but under electrical control by the A/C's PRIM computer controls - which I would call - following the documentation - "autotrim".

The THS is a major control surface,.... in Normal Law it is taken care of automatically, but in Alt Law the situation changes. In Alternate Law, the "autotrim" is still active. See page 75 of BEA Report (English version), second paragraph of Phase 3 Section, in two sentences: "THS began a movement... It should be noted that in Alternate Law the auto trim is still active."


As for the THS doing its 3°NU to 13.6°NU thing; it had nothing to do with 'autotrim'. It was operating in an integrated demand/time manner to neutralize the elevator input demand - nothing more.
The NU move from -3 to -13 degrees was controlled electrically, as the Autotrim was active - see the BEA Report, same section, and more.

You're using the word "neutralize", while some documentation is using the word "compensate".

As the THS's role is in stabilizing the pitch attitude that has been reached by means of the Elevators - that's why the name - perhaps "compensate" is closer to the definition of the function? although I can see how "neutralize" can be used with the same meaning in mind.

Once the THS is in a NU at a certain angle, a further Elavator NU at a certain angle is summing the NU effects of the two, while if the Elevator goes ND at a certain angle, the effect is a difference as long as the THS is at an opposite (pitch, or angle) position.

The elevator demand was wilful. It is generally accepted that the CVR transcript shows confusion, which puts a certain doubt if such a demand was in full awareness of the situation, and with a clear defined goal.

The A310 TAROM approach zoom climb/stall/recovery (http://www.bea.aero/docspa/1994/yr-a940924a/pdf/yr-a940924a.pdf) incident at Orly on 24 September 1994 provides some insight into the movements of the THS. ..... Also, the incident was marked by NO stall warning, and the reason given was that AoA SW thresh-hold had not been reached before the IAS had dropped below 60 KTS and the control logic inhibited it. Sound familiar?I am familiar with the Tarom case, as well as some other successful Airbus Stall Recoveries, as I've read several reports, watched 3D renditions, but a while ago.... For the Tarom case, the BEA English version of the report, which you've pointed to, has the CVR transcript in a mix of French and English - with several instances of "alarme decrochage-cricket" for "Stall Warning" at page 54.

Despite their mistakes, they've shown their airmanship, and pilot talent, achieving quite a recovery from only 4100ft to 800ft (250m), 60 degree AOA and 30 knots Stall, landing shortly after.

Essentially the same logic with minor changes has been around for a long time. Unfortunately it requires incidents/accidents before everyone realizes how some of this control logic works.That's so true...

[HarryMann]

Stick pushers are not there to deal with crew unaware of what goes on around them. They are potentially very dangerous devices and are fitted only to aeroplane that are considered to be more dangerous without then with them. DP Davies has written an excellent treatise on stick-pushers, pages 130-140. His analysis of risks and benefits connected with pushers is amazingly applicable to any other envelope protection too.Yes, they are (potentially dangerous), which is no doubt why RRR mooted a stick shaker, and I (knowing someone would bite) extended the argument to a stick shaker... in a specific context.. that is where a plane can easily be held in a stall by the crew with a neutrally stable trim - AF447

Also bear in mind that D. P Davies was writing quite a while ago now... technologically speaking.

[JenCluse]

Thank you, A33Zab, for that detailed info on the ISIS. My reservations are disarmed, with just one small niggle: one wonders why yet another generation of standby instruments don't maintain an internal backup power source, good for 10, but ideally 15-20 mins of self sufficiency.

But as the question relates to AF447, this little device should have been more than adequate to coarsely validate any suspicions or doubts concerning the primary displays.

Thanks again.

[DozyWannabe]

Also bear in mind that D. P Davies was writing quite a while ago now... technologically speaking.

Yes, and he made a weary aside in the book that some pilots were complaining about the presence of stick-pushers as potentially unsafe and unnecessary technology usurping their authority.

Then, a month after the final edition of the book was published a Trident crew comprised of a senior Captain and two relatively junior co-pilots (both S/O) on climb out from LHR retracted the droops too low and slow, approached stall and inhibited the stick pusher. The result was a full deep T-tail stall, pancaking into a field in Staines and 118 people dead.

[EDIT : On a quick fact-checking browse, I decided to have a look at Wikipedia's entry on BEA548/G-ARPI, and I have to say it's probably one of the most impressively researched articles I've seen on the site.

Now, superficially there are at least as many glaring differences as there are similarities between AF447 and BEA548 - the command gradient was very similar, and both incidents involved a loss of control following inappropriate handling, however the former was cruising, the latter was on climb-out, and while the Trident was probably the most advanced airliner in service at the time, the A330 is something else entirely. However what struck me when looking at the Flight articles in the wake of BEA548 was how similar the questions being asked and the debate over automation and warning systems is.

Check this out (apologies for the C&P, but I'm truly gobsmacked by the similarities - emphasis mine):

From 1972 | 3235 | Flight Archive (http://www.flightglobal.com/pdfarchive/view/1972/1972%20-%203235.html)

Captain Evans agreed that some of the line captains were unhappy about a situation which meant that they had their least-experienced crew member in the right-hand seat and the court heard a particularly strong adverse comment from one captain, made coincidentally only a few hours before the PI accident. "It must be remembered," said Capt Evans, "that the P2-only pilot in the P2 seat was trained as a P2 to exactly the same standard as the pilot who would be new to line operation but who had been cleared both as a P2 and a P3."

Mr Thomas referred to the sequence of warnings which would have illuminated shortly after the operation of the droop lever on the accident flight ... and asked Capt Evans whether he would agree with a remark made earlier by the Commissioner, Mr Justice Lane, that it would be very unlikely that a pilot in this particular circumstance would notice the amber
lights: "Everything would happen so quickly that the stick push would probably be the one thing which would be occupying his attention." "Yes," replied the witness, "I think I would agree with that, the big problem here being that of recognition."

Capt Evans said that he had flown through the accident sequences on both the BEA Trident simulator and the Hatfield rig. "Did you notice the droop lights and the amber lights in that period before the stick shake?" "They
were there but not very meaningful."

...

Spurious shakes and pushes were next considered against the background of a suggestion that Trident crews had become conditioned to regard any operation of either stick shake or push as false. "Do you think that pilots generally had come to regard the stall warning system as one that was liable from time to time to operate falsely?"
"One that was liable to operate falsely from time to time, yes." "One that they may have distrusted?" "I do not think so," replied the witness.

Here's one for the force-feedback/straight to Direct/Manual Trim supporters to think about, also note the trim situation :

The fact that PI was carrying nose-up trim during the accident sequence because the autopilot had been trimming to a speed lower than that which should normally have been demanded was discussed, and Capt Evans said that when he had been in the simulators he had been surprised
at the stick-force required to hold the stick forward. "I found that one held the stick forward in order to increase speed and once it had built up to something of the order of 230 to 240kt, bearing in mind that the aeroplane was trimmed to about 160kt, the pressure on the control column was quite considerable; it was difficult for me to leave one hand available to trim the aircraft out at that point." On his first run on the Hatfield rig, when the pusher was dumped at the third cycle as in the accident, Capt Evans
had not been able to recover the aircraft in the height available. "Was this partly because you found that to get the stick forward required more force than you had expected?" asked Mr Thomas. "It certainly required considerable
force for me," said Capt Evans, adding that he had needed to use both hands.

And from 1973 | 1189 | Flight Archive (http://www.flightglobal.com/pdfarchive/view/1973/1973%20-%201189.html) :

The extent of the dilemma in trying to establish the actions on the flight deck, and the motives for them, is thrown into harsh relief by one paragraph in the report. Discussing the possibility of recovery even as late as the point of the dumping of the pusher, Mr Lane says: "Unfortunately, that way of expressing the matter begs the question. If one could have asked the crew at second 127 why they were not flying a recovery, they would surely have said 'a recovery from what?' "

Plus ça change indeed! ]

[airtren]

Clandestino,

This succinct rephrasing of the sentence I asked about reads and means different, as you are well aware.

In the rephrasing, your wording "overpowering" is an element that is at the core of the reasoning behind why to have, or not to have an automatic move NU of the THS during Stall, when Elevator action need be immediately fully efficient.

I should clarify that I was looking for your own words, and your own supporting explanation, as short statements alone are often too brief to be a platform for a discussion, and so can be a book reference.

This last book reference is illustrative - you may find an appropriate appreciation. Unfortunately, I don't have the technical means to share the collection of excellent documentation on theoretical and practical aspects of the workings of the THS and Elevators which I have, for being able to reciprocate.


Yes.

Trimmable horizontal stabilizer, on any such equipped aeroplane, is always the most powerful control surface and easily overpowers the elevator.

Get this book: [pic]
and read from page 35 onwards for details.

[USMCProbe]

AlphaZuluRomeo;

Thanks for the info I didn't know that the FPV would go away, but it makes some sense, or at least did to some design engineers.

I do have a few questions about this. If the FPV gets "kicked out" by the system, how long does it take to come back?

The constraints that you wrote were that one of those parameters hat to have been met, by ALL ADRs, at the same time. The 60 knot parameter should have only been met spuriously, and transiently. The 20,000 fpm would have also been transient only, as the average was less than 15000 fpm.

Was there any indication about how long the FPV was unavailable?

Thanks

[mm43]

Hello airtren;

From BEA Final Report into TAROM A310 incident:-2.3.2 Analysis of the Stall Protection Logic
Stall protection is organized around three angle of attack thresholds, that for Alphafloor, that for Alpha-trim and that for triggering the stall warning (see 1.16.1.4).

Alpha-floor protection could not play its role as, when angle of attack of 14.5° was reached, the throttle levers were already on maximum thrust.

Alpha-trim protection was triggered at a value for angle of attack of slightly less than 15° in conditions where the flight dynamics were close to the extreme. It should be noted that it also functioned after coming out of the stall by giving the opposite order
to the THS.

The stall warning did not sound and the stick shaker did not operate in the flight phase prior to the stall. When questioned, the aircraft manufacturer indicated that the cause for non-operation of these two warnings was the disturbance of the angle of attack sensors due to the dynamics of the aircraft’s movements, with the speed having dropped below 60 kt before the angle of attack reached 17.5°. The flight crew had, however, been warned of the approach of a stall by buffeting.Back to AF447. Semantics can always be an issue. Neutralizing the elevator demand means just that, i.e. if continuous demand is made either NU/ND the THS will move in that direction until such time as the SS is placed in the neutral position. In Alt Law the Alpha protections are not available, and for this reason the elevator demand becomes a THS command as explained. At no time was the SS placed in the neutral position which would have enabled the autotrim function to maintain 1g, so effectively that function was over-ridden by the PF. Hence my reason for saying that 'autotrim' had nothing to do with it.

As an example, go back and have a look at the the initial zoom climb. The initial elevator NU commands were not aided by the THS moving because the allowable 'g' in Alt Law was exceeded ([+1.25/-0.75] where did I get that? Don't know - must have read it somewhere). During the climb the THS moved to maintain the pitch attitude when the SS movements were nominally around the neutral position, but when the 'g' went negative it moved back to 3°NU and only started tracking toward maximum with continued SS NU as the aircraft proceeded to leave the flight envelope.
http://oi52.tinypic.com/2itkgib.jpg
The red overlaid lines on the Normal Acceleration and Elevator traces are to highlight the similarities, while the THS is also similar but initially with little movement.

As usual, you or others may disagree and I am open to other interpretations.

[RetiredF4]

mm43
Back to AF447. Semantics can always be an issue. Neutralizing the elevator demand means just that, i.e. if continuous demand is made either NU/ND the THS will move in that direction until such time as the SS is placed in the neutral position. In Alt Law the Alpha protections are not available, and for this reason the elevator demand becomes a THS command as explained. At no time was the SS placed in the neutral position which would have enabled the autotrim function to maintain 1g, so effectively that function was over-ridden by the PF. Hence my reason for saying that 'autotrim' had nothing to do with it.

It has nothing to do with semantics, some of your statements are wrong. Airtren explained it already, let me try it again. Others feel free to correct me.
First some references out of LTTM (Technical training manual)

GENERAL
The pitch control is achieved by the two elevators and the THS via the computers, and controlled by the pitch side sticks orders or autopilot commands.
Max elevator deflection :
30° Nose up
15° Nose down.
Max THS deflection:
14° nose up (THS)
2° nose down.

In AP mode, the Flight Management, Guidance and Envelope Computers
(FMGEC)
send the command orders to the FCPCs ; the FCPCs transmit them to the FCSCs

THS General
The aircraft has a Trimmable Horizontal Stabilizer (THS) , which has two elevators, for pitch trim control. The two elevators are attached to the trailing edge of the THS. The THS is attached to the rear fuselage and moves about an axis to permit pitch trim. The hydromechanical operation system of the THS (referredto as THS actuator) is controlled electrically by the Flight Control Primary Computers (FCPC) and mechanically.

Operation/Control and Indicating

There are three control modes for the THS:

autoflight (electrical control)

manual (electrical control)

standby (mechanical control)

In the autoflight mode the command signals fromthe autopilot are sent to the
FCPCs. The FCPCs transmit autotrim signals tothe electric motors which control the THS actuator.

In the manual mode the command signals from the side sticks are sent to the
FCPCs. The FCPCs transmit autotrim signals to the electric motors which control the THS actuator. The computers elaborate command orders to the servocontrols, depending on the different control laws.

In the standby mode the command signals are transmitted mechanically from
the control wheels to the override mechanism. The override mechanism cancels the autotrim signals from the FCPCs. It transmits the mechanical command signals directly to the hydraulic motors of the THS actuator.


ALTERNATE LAW WITHOUT PROTECTION
In this case, the pitch protections are lost except the load factor protection.
This alternate law without protection is activated in the FCPCs after a triple
ADR failure.

DIRECT LAW
In pitch Direct Law, all the pitch protections are lost.
The elevator deflection is proportional to stick deflection.
The autotrim function is lost and only the manual control of the THS is available.

Laws Reconfiguration - General
The reconfiguration of control laws is different in pitch axis and in lateral axis.
Control law reconfigurations are divided into two families :
- ALTERNATE
- DIRECT

In the event of loss of the normal control laws:
When the conditions required for keeping the normal control laws are no longer fulfilled, the control laws are reconfigured. The various degraded law states possible are (in flight or upon flare):

Roll and yaw:
- Yaw alternate law


Pitch:
- Nz law (with limited pitch rate and gains)
- Vc PROT law
- VMO2 law
- Pitch direct law

The laws called ”Alternate” are engaged when the protections related to the
normal laws (ALPHA 1, VM01) are lost.
The laws called ”Direct” are engaged when the Nz law is lost.

Pitch
The aircraft pitch control is achieved from the side sticks and in certain cases, from the pitch trim control wheels, which act on the elevators and on the THS, depending on the different laws.


Nz law
This law, elaborated in the FCPCs, is the normal pitch law engaged in the
flight phase.
Through a pitch action on the side stick, the pilot commands a load factor ;
the Nz law achieves this command, depending on the aircraft feedbacks, so
that:
- The short-term orders are achieved by the elevator servo controls.
- The long-term orders are achieved by the THS actuator (autotrim function).
The gains depend on the Vc, on the flap and slat position and on the CG
location.
In addition, the Nz law permits to achieve:
- A load factor limitation, depending on the flap and slat position.
- A bank angle compensation, for bank angles lower than 33°.
- A deflection limitation of the THS in the nose-up direction in the event of
the activation of the high angle-of-attackprotection, the excessive load
factor and the excessive bank angle exceeding.
The Nz law is such that the aircraft response is quasi independent of the
aircraft speed, weight, and CG location. If both ADIRUs are failed, the Nz
law is kept, but with limited pitch rate and gains. A consolidation of the vertical acceleration and pitch attitude rate is then performed via the two accelerometer units.

TURBULENCE DAMPING FUNCTION
General
The purpose of the Turbulence Damping Function implemented in the Electrical
Flight Control System is to damp the structural modes induced by atmospheric
turbulence.
Architecture
The Turbulence Damping Function consists of two lanes:

Longitudinal lane
The longitudinal Turbulence Damping command is computed by the FCPC1
(FCPC2 as a redundancy) as a function of the Nz accelerometer information.
It is added to the normal law command and transmitted to the associated elevator servo-controls.

Rear lateral lane
The rear lateral Turbulence Damping command is computed by the FCPC1
(FCPC3 as a redundancy) as a function of the informationof a specific Ny
accelerometer located at the rear bulkhead level.
It is added to the normal law command and transmitted to the associated yaw damper.
Specific equipment
The equipment specific to the Turbulence Damping Function are:
- the TURB. DAMP pushbutton switch
- the Ny front accelerometer
- the Ny rear accelerometer.

In the standby mode the command signals are transmitted mechanically from
the control wheels to the override mechanism. The override mechanism cancels the autotrim signals from the FCPCs. It transmits the mechanical command signals directly to the hydraulic motors of the THS actuator.



As we know from BEA, Alt2b was active, which means concerning the pitch that we lost some protections (Alpha1, VCprot..), the rest stays basically the same.

In layman term the SS commands loadfactor to the FCPC, where this demand is transfered to a n elevator deflection to achieve this demand. SS position does not represent elevator position. As can be seen in the FDR traces the elevators follow the computer demand, in our case the elevators stayed in the full NU position also when SS was not full up That´s because the computers tried to achieve the loadfactor demand. The THS was already full up and the aircraft could not achieve the desired loadfactor with full elevator NU. Later with relaxing of NU command and even with some ND command the reduction of the elevator from full NU was only by 15° to 15° NU, because that was enough to achieve the new corrected loadfactor demand. As the elevators where deflected NU, the trim command to the THS at that time would still have been NU instead of the necessary ND.
Only when the SS input would have been held ND long enough to change the loadfactor demand significantly (dont know if loadfactor protection would come into play) and the elevators would need a ND deflection to achieve this new load factor demand, then the THS trim would start to wind the THS down to neutralize the elevator position.

Why did the THS first not move, and later on continously full up? First the elevator authority was enough to execute the loadfactor demand with little deflection only, with decreasing speed the Computers ordered the elevators to further NU to follow the demand, and then the THS started to compensate for the elevator deflection. But the still decreasing speed and still present loadfactor demand led to a continuos NU command of the elevators and the THS in the following time frame.

To sum it up in my short words, pitch control in Alt2 is the same as in normal law with AP off, except vital protections lost. No change of Autotrim. The mainly NU input of PF ordered a loadfactor, which the computers could not achieve with elevators and autotrimming the THS in the decreasing speed environment. The few moments of reducing NU order or even giving ND order did not bring the elevator into the ND command region, therefore THS stayed all the way NU.


mm43
As an example, go back and have a look at the the initial zoom climb. The initial elevator NU commands were not aided by the THS moving because the allowable 'g' in Alt Law was exceeded ([+1.25/-0.75] where did I get that? Don't know - must have read it somewhere). During the climb the THS moved to maintain the pitch attitude when the SS movements were nominally around the neutral position, but when the 'g' went negative it moved back to 3°NU and only started tracking toward maximum with continued SS NU as the aircraft proceeded to leave the flight envelope.

I couldn´t find the reference to this loadfactor protection either, although i remember some similar postings here that it was not active due to the nature of the speed failure? A

BEA in its report mentiones the ALT2B law as present, but does not describe in detail, what kind of protections had been lost and which ones still had been active. In my references i couldn´t find what the letter "B" in Alt2B stands for.

Would be interesting to know, how much unloading in term of g would have been accepted by those remaining protections. Because that again would influence elevator position and therefore autotrimming. Also if the turbulance dampening mode was activated and if that one would influence the response to inputs.

IMHO direct law from beginning might have led to an different outcome.

Sorry for the long post.

[henra]

Non sense - Why would you pull the sidestick half way ?


That's the 1mio$ question.
And it effectively applies in this case.
Why did the PF pull half way NU ?

As others already said: In a non FBW aircraft at lower Alt it would have probably shed its feathers or wings (probably both) in much less than 5s after such a pull on the controls. He probably wouldn't have had enough time to sense the feedback and react before being reduced to a lawn dart.

[mm43]

RetiredF4,

Thanks for the comprehensive response. Will analyze it in the morning and respond.

[GerardC]

@ A33Zab. Thank you for your post :
GerardC :after the, so far unexplained, FBW malfunction

A33Zab : Nothing to explain here, there wasn't a FBW malfunction.Sorry, I have not been clear enough : at 02:10:06, AP disconnection was unexplained to the crew. (No ECAM message pointing clearly to the cause of this disconnection until 02:12:XX).
Isnt'it reasonable to assume they diagnosed some sort of "computer malfunction" ?

GerardC : and smarter design are solutions for a better flight safety

A33Zab : but what if a smart design is not understood and not properly used by the user? like AF477.
Maybe "users" need better real time information from the system to "understand" what the "smart design" is doing.

IMHO there is room for improvement in the way AP disconnects in those Airbii : AP should not be allowed to quit before a clear ECAM message pointing to the reason for disconnection is displayed.

Once again : "why is such a smart AP/FD system unable to maintain for a few seconds the average pitch of, say, the past 5 or 10", until the proper ECAM alert message is displayed ?"

[henra]

BEA in its report mentiones the ALT2B law as present, but does not describe in detail, what kind of protections had been lost and which ones still had been active. In my references i couldn´t find what the letter "B" in Alt2B stands for.



According to
http://www.smartcockpit.com/data/pdfs/plane/airbus/A330/instructor/A_0-Flight_Laws.pdf

the loss of Low speed stability and High speed stability depend on the ADR's.
With dual ADR failure Low speed stability is lost, with triple ADR failure High speed stability is lost.

What is not quite clear to me is the question wether and if so how quickly these protections resume once ADR failure has been recovered.
With the speeds being around the 60kts threshold it would be interesting what that means for these speed stability functions.

[rudderrudderrat]

Hi GerardC,

why is such a smart AP/FD system unable to maintain for a few seconds the average pitch of, say, the past 5 or 10", until the proper ECAM alert message is displayed ?
I would go further and ask why it could not simply remain in some "ATT"itude where it would maintain pitch and bank as selected by the pilot with no protections. (CWS)

It would have given them more time to comprehend their status without having to struggle to maintain wings level. (I've still not flown in ALT Law at cruise FLs despite 7 years on A320s - and I don't think this crew had either)

[JD-EE]

RetiredF4 & airtren, if I wish to be provocative I can argue that AF447 suggests we have gone as far as we can go on transport aircraft which are not fully automated with no humans involved on the flight deck except as special deluxe SLF seats with all its controls disabled, permanently under computer control. That is an "improvement" in the current direction FBW is moving. {o.o}

At least with AF447, data available on the plane, and some improved computers and algorithms the pilots didn't even need to know there was a loss of airspeed indication from all three pitot tubes. There probably are other sorts of incidents that would benefit from a human pilot or two in the cockpit. But, if a rule can be evolved for the humans to follow, wouldn't a computer follow that rule better?

Humans are for when the rule based flying vanishes. But, the sense I get from descriptions of flight training as have floated through this discussion is that it is very very rule based with a lot of if-then-else-endif involved. That is a computer's playground. And on a computer's playground humans do very badly. While you add a three digit number in your head the computer has added millions or billions of them (depending on whether you are autistic or not.)

With that in mind, I am tempted to hand in my indictment for ABI, the airlines, and the regulators for trying to convert humans into computers following rules rather than having the computers follow the same rules with far better instrument scans for instruments the humans don't have a habit of scanning. Rule based actions are not the best use for humans.

[JD-EE]

Clandestino, you don't seem to get it about communications.

Both pilots, PF and PNF need to start the communications ball rolling, "ATR/AP Disconnect. ALT 2". PF starts to control the plane. States what he is doing. (NU??) The PNF breaks in right there with the correct drill - control roll maintain pitch maintain thrust. If PF does not respond, then he really does need to be hit with a 2 day old dessicated pizza across the face to get his attention. A small inflatable beach ball would probably do.

You cannot have a team if the members never talk to each other so they know what they are doing unless they've worked together so long they KNOW almost instinctively what the other is going to do. SEAL team members may be at that peak of physical and mental condition to achieve this. Mortals like us have bad days. Chatter keeps the other person informed and offers the brain check when something breaks down.

I commented further about the automation above.

[JD-EE]

Clandestino asked, "Just when did he relax the backpressure? You are acquainted with sidestick traces from 3rd intermediate report, aren't you? A couple of times stick goes forward but quickly gets behind neutral again and stays there before elevators even have the chance to get to neutral."

Page 108 English report shows between 02:12:33 and 02:13:05 the elevator moved from the stop, -30, up to a peak of -15 and from 02:13:35 until 02:14:20 it again staggered towards neutral getting to a peak of -15, again. Those more or less agree with the stick inputs. (The times are approximate given the sizeof the graphs and my laziness.)

(Just providing times for you.)

[JD-EE]

GerardC, I am not aware of any place in the BEA version of the FDR timelines where the ECAM messages are shown. Is there one? Or are you using ACARS times for those? Remember the ACARS reports are significantly delayed from what appears in the cockpit.

[AlphaZuluRomeo]

@ airtren : You're welcome, Sir.

Please note the THS logic in normal law (from the FCOM):
Automatic pitch trim is frozen in the following cases:
- manual trim order
- radio altitude below 100ft for flare
- load factor lower than 0.5 g
- in high speed protection
When angle of attack protection is active, THS is limited between setting at entry in protection and 2° nose down (i.e. further nose up trim cannot be applied).
Similarly, when the load factor is higher than 1.3 g, or when the bank angle gets outside +/- 33°, the THS is limited between the actual setting and 2° nose down.
NB : 2° ND is max position ND of the THS.
So basically, the idea I proposed is already the way Airbus'logic deals with the THS in Normal law. Let's do the same in Alternate law.
I do agree it won't have helped in other cases (i.e. Perpignan D-AXLA), but for the latter, 10" after the stall warning the aircraft go to Direct law for ~25" (then Alt law for the last 25" of flight) ; let's remind that in Direct law, there is no auto-trim (USE MAN PITCH TRIM) = no return to neutral for those 25" either, if we follow your proposed logic, then no help either.

ROLL is DIRECT, RUDDER is DIRECT, but PITCH is NOT DIRECT, and it is UNPROTECTED.
Wrong, think g protection, still active in ALT2 (pitch is not direct, but it is protected)
Also, yaw (rudder) is not direct in ALT2 either, damping & compensation are provided. See the FCOM.

If the TRIM is active, it absolutely compromises the elevators' authority.
Not under 180kt.
Or are you advocating for a trim free aircraft? I agree with Clandestino on #181.

If ACTIVE, it is also sluggish (behind), and the Pilot must form a different response to variable combinations of TRIM/ELEVATOR, which he cannot see, feel, or "hear".
Wrong. See my previous post. One can see where the THS is.


@AlphaZuluRomeo,

The IAS was grossly under reading. The acute angle of attack to the pitot tubes rendered them useless. The fact that the aircraft is airborne should be sufficient logic to maintain stall warnings and Angle of Attack.
Yes, I'm aware of that, airspeed was more in the 100kt range.
But the AoA remains unreliable under a (real) airspeed of 60kt. Then isn't having movable pitot probes a better solution than to remove the 60kt limit? Such probes exists, see the Rafale for example (one probe for AoA & pitot).

[RetiredF4]

JD-EE

RetiredF4 & airtren, if I wish to be provocative I can argue that AF447 suggests we have gone as far as we can go on transport aircraft which are not fully automated with no humans involved on the flight deck except as special deluxe SLF seats with all its controls disabled, permanently under computer control. That is an "improvement" in the current direction FBW is moving. {o.o}

At least with AF447, data available on the plane, and some improved computers and algorithms the pilots didn't even need to know there was a loss of airspeed indication from all three pitot tubes. There probably are other sorts of incidents that would benefit from a human pilot or two in the cockpit. But, if a rule can be evolved for the humans to follow, wouldn't a computer follow that rule better?

Humans are for when the rule based flying vanishes. But, the sense I get from descriptions of flight training as have floated through this discussion is that it is very very rule based with a lot of if-then-else-endif involved. That is a computer's playground. And on a computer's playground humans do very badly. While you add a three digit number in your head the computer has added millions or billions of them (depending on whether you are autistic or not.)


I unfortunately must agree with your assesment.
There was a time, when pilots used their knowledge (f.e. about aerodynamics, systems, navigation, flightphysiological aspects, weather, ATC, and so on ) to plan and execute the flight. There was a lot of self learned and self aquainted behaviour envolved, sometimes taught as "technique". That was hard work, left a lot room for errors and caused together also some undesired accidents.

New equipment together with automation reduced the workload significantly. To improve safety further, crew reaction to normal and abnormal procedures was developped and streamlined in think tanks from the manufacturer in relation to the new automated systems, and finally we´ve got what we have now and as you describe it.

Unfortunately the knowledge diminished or in some areas got lost in this process as well, as it seemed to be not needed any more. Look at the FCOM or FTCOM, it is a quarter the size of my old F4 Phantom aircraft Dash One. We aditionally had handouts about aerodynamics, navigation, WX, Radar operation, just to name a few. It got replaced by SOP and ECAM and QRH with step by step processes, which can be done by anybody who can read and operate the keyboard of a computer.

There seems to be no longer enough background knowledge available to understand, why those step by step procedures have to be followed and applied and what kind of reactions the application of those procedures will produce in the airframe.

There are for sure pilots out there and especially present in this forum, who still care and try to stay ahead of those procedures and not become the slave of them, but to be the knowlegable executioner of those procedures. They read accident reports, they talk with technicians and engineers, they ask questions and they make up their mind. And when they might find themselves in a similar situation like AF447, they will have learned from the case before the report is out and before the procedures had been changed. They would use manual trim to get the THS down although it is nowhere written in the procedures.

But how many of the pilots are participating here, how many are reading here (or in other similar forums), and how many just go home after their flight and call it a day? What are operaters doing to keep their pilots up to the notch, even improve their knowledge base except order the standard sim sessions? What do the regulators do to control and improve the knowledge after handing out the licence except to manage the renewal of the licence?

Its the system, that is sick and needs treatment badly.

[AlphaZuluRomeo]

What is not quite clear to me is the question wether and if so how quickly these protections resume once ADR failure has been recovered.
They don't resume, Alt law is latched for the rest of the flight if applicable for more than a short period of time (~10" IIRC).
Only if the ADR failure is really transient would the aircraft revert back to Normal law (and associated protections).

I do have a few questions about this. If the FPV gets "kicked out" by the system, how long does it take to come back?
According to the report, as soon as data needed are back available & limiting contraints no more met.

Was there any indication about how long the FPV was unavailable?
I wonder if one can guess that from the FDR traces. Perhaps with the FD availability?
However:
- The speed as NCD (no computed data / under 60kt) was more than transient => see § 1.16.6.1 of the 3rd report.
- The FPV was asked only shorly by the crew (during the end of the 2:11 minute IIRC), that's in the report.

[Clandestino]

Also bear in mind that D. P Davies was writing quite a while ago now... technologically speaking.I do and that's why consider chapters on fuel dippers, braking parachutes etc. as mere historical curiosities. I'll assume that you are well acquainted with the book so I would really like to know which parts of it did you find to be made obsolete by technological progress? I couldn't find many. IMHO lessons that should have been learnt from mr Davies, yet sadly were not, too often play a part in incident and accident reports in our day & age.

I was looking for your own words, and your own supporting explanation, as short statements alone are often too brief to be a platform for a discussion, and so can be a book referenceDiscussion with people proposing improvement of the system of which basics they are unaware and unwilling to learn, can be productive only by pure chance.
Opinions are debatable, facts shouldn't be so.

Unfortunately, I don't have the technical means to share the collection of excellent documentation on theoretical and practical aspects of the workings of the THS and Elevators which I have, for being able to reciprocate.'Tis a pity. I am sure that treatise on THSes that supports the notion of automatically zeroing them in case of stall warning should be very interesting read.

[+1.25/-0.75] where did I get that? Don't know - must have read it somewhere +2.5/-1.0 clean. 1.25 would quickly rob you of pitch authority, especially when banked.

I bet they would have felt it.

Instrument rated pilots tend to fly according to instruments, not according to their senses. It is not just because regulations demand so, it also greatly increases their lifetime expectancy. Properly trained IR pilots are even able to reconstruct the whole picture of their position and direction even when some details of it are missing. We call it "partial panel flying" and need for it was absolutely not abolished by the advent of glass cockpits. Definition of "properly trained IR pilot" includes, but is not limited to: proper initial IR training, proper type rating training and proper recurrent training.

Clandestino, you don't seem to get it about communications.

Both pilots, PF and PNF need to start the communications ball rolling, "ATR/AP Disconnect. ALT 2". PF starts to control the plane. States what he is doing. I am not against communication in principle. That lives of people taking flight often depend on clear communication in cockpit shouldn't be debatable. Where we differ is our opinion on AF447 pilot sitting in RH seat. You think he knew what he wanted to do but his actions were inappropriate for the situation and that's what I really doubt for the time being. It seems to me he was utterly and hopelessly lost and what puzzles me even more is that loss of SA spread like contagion to other two pilots.

Indignant cries of "why can't the autopilot save us when we don't know what we're doing?" remind me of certain young first officer:

I can not bear to look at the instruments as Hughen is obliged to do. Their readings are bringing me very close to panic. Where is that rambunctious youth intrigued with the essence of danger? What is happening so very quickly to the young man who thought the present world suffered from oversecurity and produced only mice-hearted men? Captain Hughen, this is all an anachronism. We belong to modern world and should therefore be secure. Please arrange a remedy for this grievous mistake at once.

[rudderrudderrat]

Hi Clanddestino,

Instrument rated pilots tend to fly according to instruments, not according to their senses.
I disagree. Do you believe that pilots ignore any sense of delta g? Why don't Boeing pilot's pull their wings off?

Some crew's response to a TCAS RA in the simulator (where there is no change in sensed delta g) can be "over enthusiastic". In normal line flying, pilot pitch rate of change is adjusted according to sensed g.

When faced with an abundance of visual and audio information, the overloaded brain will only process selected bits at any one time. The sensed vertical acceleration would direct attention to pitch. It's part of the human behaviour feed back loop.

edit.
Definition of "properly trained IR pilot" includes, but is not limited to: proper initial IR training, proper type rating training and proper recurrent training. The last two are now done in a simulator which maintains its physical heading, and can only simulate long term horizontal accelerations or very short term vertical (e.g. light turbulence). There is no spatial disorientation in a simulator. What makes you think it is now properly trained?

[OK465]

From post #201:

Also if the turbulance dampening mode was activated and if that one would influence the response to inputs.

Turbulence damping is only available in Normal Law or with the Auto-pilot engaged (and above 200 KIAS).

In ALT 2 with ADR problems the A/P cannot be (re)-engaged, hence no turbulence damping involved here.

[RetiredF4]

Thank you for the info, couldn´t find it myself.
Than can we assume that it was active before AP dropped out, as turbulence was expected?
How would that influence the FDR traces before and after AP dropout and would the AC behaviour be more rough thereafter, reflecting in the traces?

[Lyman]

AZR Unprotected as in STALL protection. 'g' limits in AL2? So, yes, the STALL protection is lost, and that is not related to 'g' protection, as we see on the climb to STALL?

-0.75/+1.25 Inhibits the THS. The time spent on either side of these values meant the THS was not trimming. When 'g' returned to ~1, the THS went straight to FULLNU, to chase speeds.

Whether the THS is inhibited by T/Damping? Dunno.

As to the visibility of THS when it is following the Elevators, (on the panel), my question is meant to address the need for the THS at all, here, and only if it is absolutely necessary, should the scan be plumped up with yet another "Where is it?"

[airtren]

Clandestino,
We don't have to exchange posts, if we can't keep it technical. I apologize in advance, if I got myself, or get in this post, side-tracked into "personal".

I was also looking for your own words, and your own logic, so that it would open up your own understanding, your own thinking, putting the apparent cliche arrogance of "I know better, you know nothing" aside, so we could talk science, ideas casually, as opposed to dismissive references to, or quotations from scripts.

This Forum can be fun, but can get boring if one takes himself too serious...



I was looking for your own words, and your own supporting explanation, as short statements alone are often too brief to be a platform for a discussion, and so can be a book reference.


Discussion with people proposing improvement of the system of which basics they are unaware and unwilling to learn, can be productive only by pure chance. Opinions are debatable, facts shouldn't be so.


No, you would not find that there, as I am not quoting a script.

Did you think about it? Did you tell yourself what is the difference in its effect at deep Stall at high altitude, between an announced THS move to an Optimal Stall Recovery position (as mentioned, tbd, but Neutral as a start) by automation, and a THS moved by means of Manual Trim by the pilot?


'Tis a pity. I am sure that treatise on THSes that supports the notion of automatically zeroing them in case of stall warning should be very interesting read.

[A33Zab]

@RetiredF4:

BEA in its report mentiones the ALT2B law as present, but does not describe in detail, what kind of protections had been lost and which ones still had been active. In my references i couldn´t find what the letter "B" in Alt2B stands for.


ALT1:
Pitch attitude (Θ) protection lost.
Hi Speed and Stall speed are alternate.

ALT1A:
As ALT1 but Stall protection is lost.

ALT2:
As ALT1 but lateral normal law is lost and replaced by
lateral alternate (Roll = DIRECT; Yaw = Alternate)

ALT2A:
As ALT2 (Stall protection is lost)

ALT2B:
Pitch attitude (Θ) protection lost.
Hi Speed and Stall protection lost.
Bank angle protection lost
---

Source: ?? my personal notes.

[airtren]

Thank you for the info, couldn´t find it myself.
Than can we assume that it was active before AP dropped out, as turbulence was expected?
How would that influence the FDR traces before and after AP dropout and would the AC behaviour be more rough thereafter, reflecting in the traces?


Good to know indeed, turbulence and its effects was an area of interest to me too.

Recently I went and re-read some sections of the BEA Report again, and I found the rereading of the mid section graphs and analysis quite useful in light of all the information I've got and reflected onto after/since the first reading.

[airtren]

Hello mm43,

Thanks for your clarifications. I understand your reference to the Stall Warning better. The Stall Warning was active later during the event, there is at least one additional reference to it at page 14 (English text), besides, the one I've mentioned in the CVR transcript (French translation).

There is another Airbus Stall recovery reference, which you may find interesting. Again, possible mistakes, but excellent airmanship, and talent to get the plane out of the Stall:
Interflug A310 Stall 1991 Sheremetyevo (http://aviation-safety.net/database/record.php?id=19910211-0)

Hello airtren;

From BEA Final Report into TAROM A310 incident:-

I think I understand your interpretation better.

At no time was the SS placed in the neutral position which would have enabled the autotrim function to maintain 1g, so effectively that function was over-ridden by the PF. Hence my reason for saying that 'autotrim' had nothing to do with it.

[airtren]

AlphaZuluRomeo, thanks for the clarification, it's very helpful.

If I understand correctly, you're suggesting "limiting any further NU move" of the THS at Stall in Alternate.

I like it - it's a rephrasing or refinement of part #1 of the earlier suggestion, with its implicit parts:

If I understand correctly - please correct me if I my understanding is wrong - there is an implicit relying on the fact that this would happen after a transition from Normal at cruise at high altitude, to Alternate, so that the THS would not be at max NU, and the Stall Warning would start well before the THS would/could get there.... I like it!.... Cruise at high altitude is an important condition...

Edit: The THS move ND is left to the initiative of the pilot, continueing to function as defined, auto trim consequence of strong and persistent Elevators ND, or manual trim THS ND.

airtren
@ airtren : You're welcome, Sir.

Please note the THS logic in normal law (from the FCOM):

NB : 2° ND is max position ND of the THS.
So basically, the idea I proposed is already the way Airbus'logic deals with the THS in Normal law. Let's do the same in Alternate law.

[Lyman]

AZR RE: "Unprotected". In STALL (Alpha), was my meaning. 'G' protection is assumed. The THS was inhibited from TRIM whilst the 'g' was out:

-0.75/+1.25. Is this not why the THS rermained at ~3 degrees on the way up? (zoom climb). Then, after 'g' re-entered -.75><1.25, the THS started to chase 'G' from inside the fence? Concurrent with STALL, of course, and that is the rub?

So, may we see it this way: that THS was dormant whilst a/c was 'maneuvering' (climbing, "bucking"), and it travelled NOSEUP MAX not only to follow elevator, but to prevent 'g' 'negative'? Perhaps a fine point, 'WHY' the THS did what it did, but there remain those two possibilities?

I think it is important to consider that though the pilot appears to be in Thrall with PIO, there is a PIO that can occur in reverse. Specifically, PIO induced 'Turbulence'? Seen it, done it.

[Clandestino]

Why don't Boeing pilot's pull their wings off?
Artificial feel (including, not limited to).

What makes you think it is now properly trained? Experience.

I was also looking for your own words, and your own logicYou needed not, sir. I strongly believe one of the world's foremost test pilots would do far better job of explaining operation of trimmable stabilizers than me.

So, may we see it this way: that THS was dormant whilst a/c was 'maneuvering' (climbing, "bucking"), and it travelled NOSEUP MAX not only to follow elevator, but to prevent 'g' 'negative'?If we ignore basic aerodynamics and even some basic physics, of course we may. However, this would absolutely preclude any understanding of unfolding of flight AF447.

[VGCM66]

MAYDAY Season 11 where available.

Episode 2: The Plane That Flew Too High

West Caribbean Airways Flight #708
McDonnell Douglas MD-82

Same thing one exception, copilot knew they were stalling.

[ChristiaanJ]

Re Alternate....
Source: ?? my personal notes.
Halfway serious question.... :
Could you have recited that list without your notes (AND clearly visualised or 'conceptualised' each particular case) in a car going down-hill with the power steering having given up, and a broken brake line - while pumping the brakes ?

In view of your valuable contributions, let me assure you: I'm NOT trying to be funny.
But as an engineer (not a pilot), I see this as unnecessary 'mode complexity'.

[VGCM66]

To RetiredF4 (http://www.pprune.org/members/302846-retiredf4)

VGCM66
Please,

if you are new to this thread and want to contribute something useful out of your expierience as an engineer or any other profession you are familiar with, it would be appropriate to read the nearly 1.000 pages filled about this flight.
That would tune you in the loop of the discussion.

There had been failures, there had been mistakes, and there are things which can be improved to reduce the probability of similar accidents.
The big question to all these matters is "why"?

Your ranting does not contribute to this task.

Please,

>Never start any e-mail with a plead, it shows what/who you really are right from the onset and it is not complementary as in not flattering believe me.

if you are new to this thread ...
>Yes, new since June first, 2009.

and want to contribute

>Not contributing just adding, hopefully and maybe another point of view.

something useful out of...

>Another opinion is always useful but I guess not in your world as some sort of supreme self-appointed commander in chief of the truth. Yours and only your truth that is.

your expierience as an engineer ...

>I am guilty of this. Twice. You?

or any other profession you are familiar with,

>Again guilty: Automatic Controls Systems and Applications. You?

it would be appropriate to read the nearly 1.000 pages filled about this flight.

>And again, I had. The very best and the very worst too. That is because PPRuNe is a free open for discussions Forum and not your Private Club playground where you have self appointed yourself as critic of posts and anything else you do not like. Not a private club of any kind and specially not yours.

That would tune you in the loop of the discussion.

>I am but it looks you are not.

There had been failures, there had been mistakes, and there are things which can be improved to reduce the probability of similar accidents.

>You can count on humans more. They'll do it again somehow and somewhere and blame something else for it at first. Ultimately, it always falls on us again.

The big question to all these matters is "why"?

>Wrong again. It is "HOW? as in: How can we stop this from ever happening again?

Your ranting does not contribute to this task.

>You only showed your supreme ignorance of a variety of different definitions as well as subjects.
> A330 will not be redesigned and humans in cockpits will be modified accordingly to stop future tragic events like flight AF447. We hope.

[airtren]

You needed not, sir. I strongly believe one of the world's foremost test pilots would do far better job of explaining operation of trimmable stabilizers than me.

I appreciate your post, and respect the deference .

[A33Zab]

Hi CJ,

Halfway serious question.... :
Could you have recited that list without your notes (AND clearly visualised or 'conceptualised' each particular case) in a car going down-hill with the power steering having given up, and a broken brake line - while pumping the brakes ?

Absolutely NOT and also NOT of any use,
it was an answer to RetiredF4s question about the 'B' in ALT2B and I mentioned
the other alternate modes and source in advance of the expected next questions.

I agree.........set and forget.

BTW, considering my car, it wouldn't even get up-hill.

[CONF iture]

Don't you adapt your inputs to achieve desired attitude, no matter what control law are you in?
Attitude is the objective, but 36 ways to reach that goal, how smooth or not you can or want to be is the director of your inputs.
Clandestino, if you start feeling Gs do you still pull harder or you just relax … ?

If you set and hold 5° and set climb power, the aeroplane will climb, after a while power available goes down with altitude, EAS goes slowly down, AoA goes gently up and aeroplane levels off when AoA reaches five degrees minus wing incidence angle.
Except that 5 degrees of AoA on AF447 and stall warning is already warning, stall itself is just about.

What about a more common sense 2.5 deg pitch ?

[HazelNuts39]

Except that 5 degrees of AoA on AF447 and stall warning is already warning, stall itself is just about.
I agree with PJ2 and others that going to 5 degrees pitch is not what one would expect a seasoned pilot to do, and is not the right thing to do in the circumstances. Depending on how quickly the airplane is rotated to 5 degrees pitch, the AoA would probably temporarily exceed the stall warning threshold of about 4 degrees, and in any case the 'stabilized' AoA in still air would move closer to the stall than with 2.5 degrees pitch. On the other hand I believe, based on 'gut feeling' rather than a numerical analysis, that if CLB power had been set and attitude had increased to but not exceeded 5 degrees pitch, that the airplane would not have stalled.

[Lyman]

So I considered just allowing a slag off of my opinion, but I think it a good question, and will pose it once more.

Per the post by airtren of the THS and Elevator traces, I notice a dormant THS, positioned at 3 degrees virtually the entire time of the climb. It then travels to 13.6 degrees Nose Up at a very consistent rate at and during the STALL WARN.

STALL Protection is lost in ALTERNATE LAW, but 'g' protection remains. The elevators work via the FCS, and respond to load, or g sense.

So Pitch is not DIRECT, and it is independent of an "active" (variable) Stick. No excessive elevator deflection to be expected. The climb shows a response to NU input, again, without TRIMMING, which is ACTIVE in this LAW.

Why then, no "compensation" for the elevator position? The stick is held back sufficiently to command autotrim, but none shows.

Why? Because the THS, its Autotrim function, is inhibited in the region above 1.25 g, and below .75 g. The g traces show that the g is consistently outside the normal Autotrim range, so no mystery. Then, as the a/c climbs to its apogee, and g re enters the range in which Autotrim is ACTIVE, the THS migrates at an even rate to its NOSEUP stop. At no time do I see a variable rate from the THS, it lumbers along from 3 degrees to the stop, seemingly independent of a reason to do so in such a manner. This all seems inexplicable to me, what is the relationship between elevator and THS in this trajectory?

That is the best wording I can come up with. I started what seemed like a fruitful exchange with MR. clandestino, but found out my question lacked aerodynamic validity, and seriously challenged basic Physics. That's it, no answer, just slag.

Oh, except for a picture of a fifty year old design, and a rejoinder that Mr. Davies knew more than he.

Fair. I think mine is a valid question, and not a steaming pile, as has been inferred. If a pile, please ignore.

[Emere]

Air France 447 - AFR447 - A detailed meteorological analysis - Comments from pilots and other aviation professionals (http://www.weathergraphics.com/tim/af447/comments.shtml)

in case you haven't seen it.

[rudderrudderrat]

Hi Lyman,

what is the relationship between elevator and THS in this trajectory?
As the speed washed off, the nose of the aircraft would want to drop naturally. The FBW computers are programmed to maintain attitude, so more nose up from the elevator is required. The stab was autotrimmed to reduce the amount of elevator deflection required.

[AlphaZuluRomeo]

@ airtren (re: #223)
Yes, you understood perfectly what I suggested :)
"implicit relying"? I wouldn't have phrased it like that, but anyway: It's a fact that the proposed solution will be of no use in a "Perpignan-like" case (reversion to direct law), nor if we take the hypothesis of "just" a reversion to alternate after the THS is already max NU.
But Perpignan was a specific case, with AoA measurement errors: one can't elaborate on that, as (reliable) AoA is required for stall warning.

As for you last §, I'm sorry I don't get your point? What do you mean?
"The limiting, which you look at, if I understand correctly, as a temporary excursion of THS control in DirectLaw could be made Stall Warning dependent, instead of a fixed time interval? Pre-defined time intervals, don't always (actually I should say they rarely) respond well to needs in very dynamic situations."

---------

@ Lyman (re: #224)
I understood "unprotected" as "not a single protection". And, as you seemed astonished that rudder was limited (NB: that's for structural limits reasons) but not the pitch, I replyied with the fact that the pitch is also limited (in g, not regarding stall/AoA, on that we agree), due to structural limits reasons.
Now, regarding the stall/AoA protection: If the aircraft could have delivered it, it would have. ALTERNATE LAW (PROT LOST) says it well, doesn't it?

[dufc]

Like others, I suspect, I have been struggling to understand the PF’s reaction(s) after he was handed control of the aircraft when the AP dropped out.

I am not a pilot but my understanding is that the PF took action(s) opposite to or at variance with those normally expected in such a situation.

The apparent lack of meaningful/productive communication between PF and PNF – not going through SOPs, etc – and the frantic requests for the Captain to return suggests to me that in essence they had no idea what was going on or where to start to resolve things.

Accordingly the PF may have decided that everything he was seeing and hearing was incorrect leading to him shutting it all out as he sought to aviate. Using his ‘gut feeling’ his default position became an effort to secure altitude and power - to give a safe “breathing space” to evaluate what was in fact happening - thus explaining NU and TOGA.

I accept that PNF may have had some (growing?) understanding but he did not communicate this with the strength required or take over control apart from a very brief period before being usurped by a PF locked into a wrong belief about what was needed almost right to the very end.

As I say, I am not a pilot but wondered if the above explanation might fit with what tragically transpired…

[RetiredF4]

A post in the thread causes myself to go back to the upset and stall and the
unsuccessful recognition and recovery from the stall.

Upset recovery training zip (http://www.mediafire.com/?jrkvp2ysl7aea25)
posted on http://www.pprune.org/tech-log/417975-new-2010-stall-recoverys-high-altitudes-8.html#post6652672

Although the posted reference is dated from August 2004, it is imho an excellent work. It is not the powerpoint presntation, but it consists of 185 pages of very interesting read.

But it deals with non FBW aircraft, has some references to them that those might not stall as being protected.

Actually if those informations and recomendations had been used from the crew of AF447, the outcome would have been most probably a different one.

I´ve copied some sentences for info.


Simulator versus reality
There are issues associated with differences between simulator training and aircraft recoveries. A simulator can provide the basic fundamentals for upset recovery, but some realities such as positive or negative g’s, startle factor, and environmental conditions are difficult or impossible to replicate. These limitations in simulation add a degree of complexity to recovery from an actual aircraft upset because the encounter can be significantly different from that experienced during simulator training. Therefore memory checklists or procedural responses performed in training may not be repeatable during an actual upset situation. The limitations of simulators at the edges of the flight envelope can also cause fidelity issues because the simulator recovery may or may not have the same response characteristics as the aircraft being flown. However, provided the alpha and beta limits are not exceeded, the initial otion responses and instrument indications of the simulator should replicate airplane responses.

The Alpha and Beta values are depicted in Appendix 3 for a lot of aircraft, from AB the A300/A310.
In short with flaps up flight validated from 0° AOA up to 12 AOA,
Wind tunnel / analythical from -5°AOA up to 12 ° AOA
Extrapolated for simulator from -5° AOA up to 30° AOA

AF447 maneuvered well outside those limits.

Startle factor
It has already been stated that airplane upsets do not occur very often and that there are multiple causes for these unpredictable events. Therefore, pilots are usually surprised or startled when an upset occurs. There can be a tendency for pilots to react before analyzing what is happening or to fixate on one indication and fail to properly diagnose the situation. Proper and sufficient training is the best solution for overcoming the startle factor. The pilot must overcome the surprise and quickly shift into analysis of what the airplane is doing and then implement the proper recovery. Gain control of the airplane and then determine and eliminate the cause of the upset.

Unloading
Airline pilots are normally uncomfortable with aggressively unloading the g forces on a large passenger airplane. They habitually work hard at being very smooth with the controls and keeping a positive 1-g force to ensure flight attendant and passenger comfort and safety. Therefore, they must overcome this inhibition when faced with having to quickly and sometimes aggressively unload the airplane to less than 1 g by pushing down elevator.

Cockpit environment
Pilots must anticipate a significantly different cockpit environment during less-than-1-g situations. They may be floating up against the seat belts and shoulder harnesses. It may be difficult to reach or use rudder pedals if they are not properly adjusted. Unsecured items such as flight kits, approach plates, or lunch trays may be flying around the cockpit. These are things that the pilot must be prepared for when recovering from an upset that involves forces less than 1-g flight.

Flight controls
Utilizing full flight control authority is not a part of routine airline flying. Pilots must be prepared to use full flight control authority if the situation warrants it. In normal conditions, flight control inputs become more effective with increased speed/ reduced angle of attack. Conversely, at speeds approaching the critical angle of attack, larger control inputs are needed for given aircraft reactions. Moreover, during certain abnormal situations (partial high lift devices, thrust reverser in flight) large or full-scale control inputs may be required. Attitude and flight path changes can be very rapid during an upset and in responding to these sorts of upset conditions, large control inputs may be necessary. It is important to guard against control reversals. There is no situation that will require rapid full-scale control deflections from one side to the other.


Stall aproach / Stall
Pilots are routinely trained to recover from approach to stalls. The recovery usually requires an increase in thrust and a relatively small reduction in pitch attitude. Therefore, it may be counterintuitive to use greater unloading control forces or to reduce thrust when recovering from a high angle of attack, especially at lower altitudes. If the airplane is stalled while already in a nosedown attitude, the pilot must still push the nose down in order to reduce the angle of attack. Altitude cannot be maintained and should be of secondary importance.

Stall recovery
A stall is an out-of-control condition, but it is recoverable. To recover from the stall, angle of attack must be reduced below the stalling angle—apply nosedown pitch control and maintain it until stall recovery. Under certain conditions, on airplanes with underwing-mounted engines, it may be necessary to reduce thrust to prevent the angle of attack from continuing to increase. If the airplane is stalled, it is necessary to first recover from the stalled condition before initiating upset recovery techniques.

Stall recovery procedure for AF447 (would have worked, imho)

Situation: Pitch attitude unintentionally more than
25 deg, nose high, and increasing.
Airspeed decreasing rapidly.
Ability to maneuver decreasing.

Nose-high, wings-level recovery:
◆ Recognize and confirm the situation.
◆ Disengage autopilot and autothrottle.
◆ Apply as much as full nosedown elevator.
◆ Use appropriate techniques:
• Roll to obtain a nosedown pitch rate.
• Reduce thrust (underwing-mounted engines).
◆ Complete the recovery:
• Approaching horizon, roll to wings level.
• Check airspeed, adjust thrust.
• Establish pitch attitude.


Pitch control
Pitch may be controlled by rolling the airplane to a bank angle that starts the nose down. The angle of bank should not normally exceed approximately 60 deg. Continuous nosedown elevator pressure will keep the wing angle of attack as low as possible, which will make the normal roll controls effective. With airspeed as low as the onset of the stick shaker, or lower, up to full deflection of the ailerons and spoilers can be used. The rolling maneuver changes the pitch rate into a turning maneuver, allowing the pitch to decrease.

An interesting read, especially IMHO for the non-flyers on this thread. Pilots should know it anyway.

[Mr Optimistic]

RF4 thanks. Only one comment: no rudder? ;)

[RetiredF4]

Mr Optimistic
RF4 thanks. Only one comment: no rudder?

I wouldn´t have used one. It was a pro spin flight control in my stall and spin training. The main emphasis of the procedure is reduction of AOA. AF447 PF concentrated on roll for too long.
The yawing and the problems of the PF to reestablish wings level flight was imho caused by using ailerons at that high AOA.

One interesting paragraph i forgot in the above post, it is from the introduction part and should have been on the top:

Note the thing about tactile feedback.

All modern jet transports are certified to exhibit adequate warning of impending stall to give the pilot opportunity to recover by decreasing the angle of attack. Whether this warning is by natural aerodynamic buffet or provided by a stick shaker or other warning devices, it warns the pilot when the angle of attack is getting close to stall. Moreover, the warning is required to be in a form other than visual. The pilot need not look at a particular instrument, gauge, or indicator. The warning is tactile: the pilot is able to feel the stall warning with enough opportunity to recover promptly. Pilots need to be especially cognizant of stall warning cues for the particular airplanes they fly. The onset of stall warning should be taken as an indication to not continue to increase the angle of attack.

[Mr Optimistic]

Given the very high AoA during the descent, it is surprising that the ailerons were effective enough for roll to be (reasonably) controlled. Not much point looking at how well teh PF controlled roll under these circumstances, at least in terms of judging his sidestick touch. Presume the ailerons were acting as drag modulating devices rather than Cl adjusters. Not so very much more and the wing would have been going backwards (AoA >90 degrees).

Interesting quotes though, so thanks again.

[airtren]

AlphaZulyRomeo,

At a re-read of your text, that last paragraph of my post is to be ignored. Originally I was not 100% if my reading of your text was correct, and at a re-read I realize that it was a misunderstanding.


@ airtren (re: #223)
....
As for you last §, I'm sorry I don't get your point? What do you mean?
...

[Lyman]

AlphaZuluRomeo

Yes, I was speaking of Alpha protections lost. I was party of a very long discussion on Rudder/Vertical Stabiliser issues long ago, and understand reasonably well the RTLU.

The PITCH is protected "IN" 'g' ? OR "BY" 'g'? Its PROT is time discrete, as it prevents transient exceedance of PITCH loads.

It demonstrably does not prevent high angles of Attack see: "the climb"

In fact, as long as the a/c rotates within a 'g' envelope, there seems to be no functional limit of PITCH UP.

Now here is the problem with that, as I see it.

447 entered the climb without a STALL protection, and STALLED. She climbed at a rapidly inceasing PITCH value, but with a controlled RATE OF CHANGE.

Wasn't PF using 'g' prot as his NORMAL LAW "stand-in"? No immediate and seat crushing PITCH UP, but a steady excursion toward the STALL AoA?

So I see "G" PROTECTION as a de facto airframe limiter, which not only prevents over load, but allows LOC, and in Losing Control, "G" PROT, having allowed it, prevents recovery "FROM IT".

How can I say that? Because the TRIM started to LOAD the airframe, in STALL, when to unload it, or allow it to be unloaded, would facilitate recovery.

Nothing Earthshattering about this, it is a point of view. Will you attempt to see it that way for purpose of understanding?

Climbing UP, PF had an airframe that appeared NORMAL, but was rotating to absurd PITCH UP. By that I mean, his climb was incremental, and indicated no disaster in the offing. (possibly).

If PILOT/PLANE communication is critical, wouldn't the a/c have annunciated to the crew[/I] that pilot input was being g managed?

In a perfect world, one then says "but he should have known this", no denying. Was there an clear alert that instructed the manual pilot that basic, and life protecting systems were operating the Plane, and not he?

I have a sinking feeling that the PF was possibly unaware "g" was molding his inputs to an incremental PITCH UP that he may have sussed was not potentially very hazardous. Otherwise, why the gruff Stick demands?

add. "Attitude" and 'g' PROT are not related. There can be (have been) circumstances where 'g' PROT might have wanted to be sent packing?

The chronic application of PITCH UP by this pilot is made easier to understand (by me), when one imagines the environment, the lack of "Response", and the inappropriate nature of some of the cueing.

THS. There is a command in medicine, that applies to the inappropriate automatic application of TRIM:

"First, do no harm"...... perhaps engineers should have a short 101 in patience at onset of disaster, as well as Physicians.

[Ian W]

Quote Clandestino:
Originally Posted by Rudderrudderrat
In ALT LAW, when he relaxed back pressure, the aircraft maintained it's nose up attitude as the speed washed off.

Just when did he relax the backpressure? You are acquainted with sidestick traces from 3rd intermediate report, aren't you? A couple of times stick goes forward but quickly gets behind neutral again and stays there before elevators even have the chance to get to neutral.

I think that this may miss the point.

If (as suggested earlier) the THS had remained at NEUTRAL rather than chasing the NU input during a stall, would the PF NU inputs have been enough with elevator alone to maintain nose up into a deep stall? Was it only the added authority of the THS that allowed the aircraft to be kept in the stall?

Perhaps with elevators alone and a THS neutral the aircraft would have dropped its nose, then (had PF kept NU) after getting flying speed its nose would come back up. Perhaps PF was expecting this nose drop 'nodding' behavior? Its what would happen in a lot of gliders and light aircraft and he was glider qualified. The lack of this nose drop possibly convinced PF that the aircraft could not be in a stall.

Was there any training or flight envelope information available to A330 crews that told of the possibility of deep stall? There seems to have been quite a lot on 'spurious' stall warnings.

So without any knowledge of deep stall, with THS having sufficient authority to hold the aircraft in a stable deep stall so no nodding ND, and previous information about spurious stall warnings: everything seems to be set up for a confused PF. :confused:

[AlphaZuluRomeo]

Lyman,

Let's go step by step:
- I do agree that the autotrim going full UP is a bad idea. Re-read my previous posts (discussion with airtren) on the subject.
- I do agree that AoA/stall (hard) protection and pitch (hard) protection were lost on AF447. That's the very raison d'être of the Alternate Law: when the plane can no more grant full protection, it reverts to Alternate (or Direct, if failures are more than what allows Alternate to be invoked).

If PILOT/PLANE communication is critical, wouldn't the a/c have annunciated to the crew that pilot input was being g managed?
You're saying - unless I'm mistaken - that "g demand" flight controls are dangerous and "unknown" of the pilot?? May I remind you that the Normal Law is also g demand? May I remind you that, apart from Dassault's Falcons, all FBW civilian aircraft use the g demand philosophy? There is no need of being in a perfect world for hoping that pilots know that!! :confused:

Was there an clear alert that instructed the manual pilot that basic, and life protecting systems were operating the Plane, and not he?
Yes there was. ALT LAW (PROT LOST) on the ECAM means just that. And this was acknowledged by the crew (see CVR transcript).
On a side note, I wouldn't call the g protection (nor the whole g demand law) "life protecting systems". They're:
- structural protecting systems for the g protection
- standart way of flying the aircraft for the g demand law.
Nor the g protection, nor the g demand controls alone will prevent any pilot to do what he wants with his plane (even stalling it, or recovering it from stall).
It will only - if needed - prevent him to do it too fast for the aircraft structure to cope with. The aircraft will retain its wings. Good idea, isn't it? :ok:

On the whole, I don't get it. What are you trying to prove? That a pilot must know its plane? I took that for granted... ;)
I do agree that a PPL being thrown in AF447's cockpit could have difficulties with the g demand law. But we're talking of a professional crew, here, with ATPL & type rating licences. :ugh:

[Zorin_75]

Was there any training or flight envelope information available to A330 crews that told of the possibility of deep stall? There seems to have been quite a lot on 'spurious' stall warnings.
Can't we put this deep stall nonsense to rest already (along with coffin's corner)?

[RR_NDB]

Hi,

As I say, I am not a pilot but wondered if the above explanation might fit with what tragically transpired… (http://www.pprune.org/tech-log/460625-af-447-thread-no-6-a-12.html#post6653272)dufc,

The "information" (we don't have it) PF received (when a/c was operating outside* it's design) led him to an "strange behavior". We need more info to understand PF persistent NU.

...suggests to me that in essence they had no idea what was going on or where to start to resolve things. When 3 qualified pilots fail, even to "understand timely", probably we have something very serious in the a/c (design, cert, training, operation, etc). The very basic info MUST be absolutely reliable and presented in real time by the a/c resources to the crew. Complex machines may fail in this aspect. And should not...


(*) F-GZCP at that night, operated "outside design":

1) Facing WX (icing)
2) Too high (briefly)
3) Stalled

[Clandestino]

Vielen Dank, Franzl, for posting "Pilot guide to aeroplane upset recovery". It is very useful reading, learning lessons from it might save someone's flying rear yet but I do not think it is applicable to AF447 case. It mentions that definition of aeroplane upset includes unintentional.In other words, the aircraft is not doing what it was commanded to and is approaching unsafe parameters. Beyond any doubt, pulling the aeroplane into stall was not intentional, however aeroplane did just what she was commanded to. Looks as if 447's crew simply forgot how to fly an aeroplane. So crashed.

Attitude is the objective, but 36 ways to reach that goal, how smooth or not you can or want to be is the director of your inputs.
Clandestino, if you start feeling Gs do you still pull harder or you just relax … ?My sincerest apologies, sir. I was under distinct impression I was debating with professional pilot, rated on FBW Airbus. I mistakenly believed that we share some common background so there would be no need for me to go into details that are ingrained in basics of airline pilots training, such as technique of instrument flying. I now see that by omitting the details I believed would be familiar to you, my answers came across as quite harsh, snobbish and disrespectful. It was not my intention. From now on I'll try to dumb it down to level of high school physics.

Most important and most often checked information in "blind flying" is attitude. If one cannot read it properly or cannot maintain it properly, the rest of instrument scan is in vain.

When pilot is handflying by sole reference to instruments, it is important to keep movement smooth and precise to avoid unnecessary maneuvering that might upset pilot's sense of balance and induce illusions of turning or banking. Pilots who, when without outside visual reference, start flying by their senses instead by their instruments, get far more often killed than not. That is known fact that is with us since there were first instrument flights and no amount of sophistication and automation is able to isolate us from it, as was correctly, in more general terms, predicted by Antoine de Saint-Exupery.

So, as instrument rated pilot, I don't fly by feeling Gs, I fly by reference to my instruments.

Except that 5 degrees of AoA on AF447 and stall warning is already warning, stall itself is just about.

what about a more common sense 2.5 deg pitch ? Non, monsieur. You are mistaking the angle of attack and pitch. Pitch is angular difference between aeroplane's longitudinal axis and horizon. Angle of attack. in the most layman's terms, is angular difference between where nose of the aeroplane is pointing and where aeroplane is travelling to through air (velocity vector), measured in aeroplane's vertical plane. So AF447 was traveling at 2.5° cruise AoA and recieved first two stall warning as it was pitching up in turbulence, they were quite short and only transient and stopped promptly. It is also important to understand that stall warning sounds before actual stall takes place. It is certification requirement, purpose of which is for pilots to have enough time to make corrective actions before actual stall takes place.

Old unreliable airspeed procedure directed pilots to fly 2.5° attitude and power from table. New procedure, valid at time of AF447 final dive, called for setting 5° pitch with climb power for a couple of seconds, while PNF takes out the table of pitch-against-power from QRH and then power and attitude are set accordingly. 2.5° pitch is fine if you know your cruise power by heart. With climb power it would likely result in overspeed.

I agree with PJ2 and others that going to 5 degrees pitch is not what one would expect a seasoned pilot to do, and is not the right thing to do in the circumstances. Depending on how quickly the airplane is rotated to 5 degrees pitch, the AoA would probably temporarily exceed the stall warning threshold of about 4 degrees, and in any case the 'stabilized' AoA in still air would move closer to the stall than with 2.5 degrees pitch. On the other hand I believe, based on 'gut feeling' rather than a numerical analysis, that if CLB power had been set and attitude had increased to but not exceeded 5 degrees pitch, that the airplane would not have stalled. I would expect them to do exactly that as UAS procedure valid at the time of crash demands it. Page 59 of accident report refers. Your gut feeling is correct, I have explained why a few pages ago.

Why then, no "compensation" for the elevator position? The stick is held back sufficiently to command autotrim, but none shows.
Demanded load factor was met by elevators alone. Reason? High dynamic pressure at start of the climb.

The FBW computers are programmed to maintain attitude, so more nose up from the elevator is required.
They maintain attitude when stick is pitch neutral. In normal and altn law they are programmed to meet G demand set by sidestick.

If PILOT/PLANE communication is critical, wouldn't the a/c have annunciated to the crew that pilot input was being g managed?One only needs to look int Flight Crew Operation Manual (FCOM) to see that info. Pilot input is always G managed, except in direct law.

There can be (have been) circumstances where 'g' PROT might have wanted to be sent packing?Only if one is purposefully set on breaking up the aeroplane in midair through overstress. Absolutely not applicable to AF447. Aeroplane never came close to G limits.


The chronic application of PITCH UP by this pilot is made easier to understand (by me), when one imagines the environment, the lack of "Response"People who don't understand that if they pull and nose doesn't come up (EDIT: sorry, I wrote that terribly wrong the first time), there are 99.99% chances they are stalled, are usually not allowed to go solo in gliders, microlights or light aeroplanes, let alone climb the ladder to professional aviation. 0.01% caters for airframe failures, control circuit failures, etc.

If (as suggested earlier) the THS had remained at NEUTRAL There is absolutely no NEUTRAL position as such on any THS fitted to any aeroplane, anywhere. There is neutral position for speed, altitude, weight, balance and configuration. It can be anywhere within stabilizer operating range.

Was there any training or flight envelope information available to A330 crews that told of the possibility of deep stall?No. It would be a lie. Airbus 330 is not prone to deep stall/ superstall/ locked-in condition.

[ChristiaanJ]

Can't we put this deep stall nonsense to rest already (along with coffin's corner)?Stop moaning about that, Zorin.....
It's a matter of terminology, further confused by the early "locked-in stall" accidents of some of the early T-tails being referred to as "deep stalls".

I would have called AF447 a "deep stall" in the general sense, in that it was a persistent, fully developped stall, with a few nasty contributing factors (no airspeed, THS, confused meatware), which lasted until FL000....

Maybe you could suggest some 'formal' terminology?