Thrust Lock
 

You raise another good point. Having read up some more on TOGA thrust lock incidents I can see that this issue alone has led to pilot overload and overspeed in the A330.
A330 experienced an uncancelable toga lock thrust condition following a predictive windshear go around
Are we reaching the point of over automation?

spatial disorientation after a stall
 

Very interesting your mention of the "spatial disorientation"

PS : But for your "AF bashing" it is not necessary ; AF has not to prove it knows how to operate planes since the 1930' ; and the Paris - Rio route is one of the oldest !

Flight Plan
 

The AF447 flight plan on the Vasquez site clearly states destination is Charles De Gaul (LFPG). So we can debunk the Toulouse destination theory if that is a reliable facsimile.

Air France 447 - AFR447 - A detailed meteorological analysis - Satellite and weather data

Barnyard Logic II
 

Lonewolf50 The real measure of AOA is degrees, not knots, of course, although everybody has been shouting for a Stall Warning below 60 knots, not a specific AOA.

I don't have the numbers, but it appears that onset of stall of the A330 at that MAC and flap is less than 20 degrees. The report shows AF447 achieved AOA in excess of 40 degrees, double the onset of stall. How much AOA does it need to measure, 90 degrees, 120 degrees? What's the point?

You have to balance Stall Warnings in extremely rare events with far more common nuisance stall warnings, in order to maintain confidence in the system. In fact there was no doubt a point in the zoom climb to stall that airspeed was near zero, and the AOA vanes would fall with gravity, probably showing a negative AOA.

Saturn,
SaturnV,
The notification from ATLANTICO to DAKAR, (contrary to what happens now), was not automatically done from one ATC to the other at that time, once the CPDLC/ADS FANS system was not fully operational at that time. According to FANS procedures, one should try to notify/(log-on) between 10 to 40 minutes before reaching the boundary. Further, there are normally two HF radios on A330's. Therefore, it was common practice to call DAKAR, at least 10 minutes before reaching the boundary on HF2, while maintaining a SELCAL watch with ATLANTICO on HF1.

But I'm with you. I'm very curious on what CVR transcripts may bring to light.

Greybeard reminds us:

...but like the weathervane, shouldn't it too be statically balanced? And if so, why would it be unreliable below 60kt? :confused:

The balance weight would not need to be out in the airstream; it could be internal, behind the aircraft skin.

Just askin'.

Static Balance
 

Every AOA sensor I've seen on ground has been full stop down (neg AOA), which indicates no static balance. Don't believe I've ever seen one in flight. :)

I don't know a lot about AOA vanes, which I why I asked for someone really knowledgeable to speak up. You can peruse pix on !!!!!!!!!!!!!! to get a good sample.

TheShadow:

Your prescience is unparalleled on this thread. We have only been provided with a sou-son of data but I believe you have nailed it. The THS situation was, I also believe, the lethal factor.

I thank you for your insights. As a current 330 pilot I will now be paying a lot more attention to its position. Especially if things ever get convoluted!

I wondered who you were. I see from a basic search you may be an ex-Viet chopper guy. How do you know about all this Airbus stuff? And especially in such detail? Please tell us more?

Your erudition is also commendable.

Please keep up this extraordinary work.

The aviation industry may have just found their spokesperson. Notice his emphasis on proper training and having AoA displayed . We should all be grateful he landed that plane safely on the Hudson and became a hero. It seems his time has come....

Sully: Training is key to avoiding air accidents - CBS News Video

True. But where that balance point is located is a matter of opinion (i.e, professional judgement). People can and do disagree. We tend to fixate on the needs of the moment and forget there is two sides to the story. Then it just becomes a game of ping-pong where we respond to one crisis by going too far to the left and another crisis by going too far to the right. It seems to me that the better course of action over the long haul is to pick a point of balance, any point, and then train around it.

I'm not sympathetic to the cacophony that wants to tweak software systems in response to every accident. The human being with their hand on the stick has to bear some responsibility, as does the entire safety systems in the corporation. Who trained this crew? Who gave them their line checks? Who oversaw the sim sessions and designed the syllabuses. :uhoh:

Edit:

I think his time has been and gone. Everything he says sounds good in theory but he fails to mention the critical factor of money. Sims are not cheap, sim sessions are not cheap, practice lessons in the plane are not cheap. How much money is a society supposed to spend in order to save 200-300 lives. Naturally, if it's your life at stake you want them to spend a lot of money. But when it's the other guy, maybe not so much. Whether it is explicit or implicit there is always a cost/benefit analysis going on. Always.

Gents,
a question - without intending to minimise the difficulty of dealing with a situation like this - espec when the EFIS appears to have been giving very confusing readings, in strong turbulence and a confused flight deck environment - and without intending to insult anyone's intelligence.
Does this model of Airbus have a standby analog artificial horizon on or near the centre of the panel as I have seen in some airliners?
In what would appear to have been a jetupset situation like this, could any experienced airline pilots viewing this thread comment on how feasible it would have been to simply set cruise power and then maintain a wings level, nose level attitude using an analog AH alone?
I know it's aviaition 101 if you go IMC in a light aircraft, but perhaps not in a heavy jet...

Artificial Horizon is an available selection, at extra cost, and was not optioned by AF for this a/c, as I understand it. I think it occupies the LHS of panel, top and left in P1's scan.

AVIATE
 

Razoray,

Part of AVIATE means don't fly through CBs when you are close to the "coffin corner". Had they deviated, all this talk of this law and that law and how to handle a stall would not be taking place. That's the real cause of this crash, nothing else. If they were already close to the CC at 35,00 ft, think of what would happen when they went to 38,000 ft.

Fuel stop
 

Soaring the Skies,

The embarrassment of making a fuel stop should never get in the way of safety. Divert if necessary then stop somewhere for fuel. That's the Captains prerogative. The Captain is totally responsible for the safety of the aircraft and it's contents. They could have stopped in the Canaries or Lisbon for fuel.

Tartare

The pressure side of the EFIS was giving erroneous information the ATTITUDE part of the PFD was still working so they had a large Artificial Horizon to enable attitude to be flown.

Bearfoil

They had a standby AH it is called the ISIS, basically a smaller version of the PFD.

So - a hypothetical question - Bearing in mind the idiosyncrasies of swept wing aircraft in stalls... if such a jet is in a fully developed stall, applying power and then using the AH/EFIS to obtain a wings level, nose level attitude would arrest the stall?

Not without a big afterburner.

Too much induced drag. Reduce AOA first, accelerate, and fly happily ever after.

Try to power out of it and everyone will be shaking their head at your airmanship as they attend the funeral.:{

Alternate destinations
 

xcitation:

The BEA Interim Report stated:

1.17.1.3 Preparation of flight AF447 on 31 May 2009
Preparation of the flight by the central flight study service
The flight was prepared between 15 h 28 and 18 h 59. Paris Orly was given as the alternate airport at destination. Given the estimated load of 37.8 t, the dossier included a main flight plan at a standard Mach of M 0.82 with an ETF at Bordeaux Mérignac with alternate at Toulouse Blagnac as well as two additional direct flight plans, one at Mach 0.82 and the other at a "slower Mach", i.e. M 0.81. A summary table of the loads offered enabled the crew to make the choice of the definitive flight plan from among these three options.

The full report is at http://www.bea.aero/docspa/2009/f-cp...90601e1.en.pdf

Lack of Stall Warning
 

To have a design where blockage of the pitots not only loses airspeed data but also (because it now shows speed is less than 60kts regardless of the truth) disables the stall warning strikes me as unwise.

Flight Plan
 

xcitation:

The BEA Interim Report stated:

1.17.1.3 Preparation of flight AF447 on 31 May 2009
Preparation of the flight by the central flight study service
The flight was prepared between 15 h 28 and 18 h 59. Paris Orly was given as the alternate airport at destination. Given the estimated load of 37.8 t, the dossier included a main flight plan at a standard Mach of M 0.82 with an ETF at Bordeaux Mérignac with alternate at Toulouse Blagnac as well as two additional direct flight plans, one at Mach 0.82 and the other at a "slower Mach", i.e. M 0.81. A summary table of the loads offered enabled the crew to make the choice of the definitive flight plan from among these three options.

The report is available at: http://www.bea.aero/docspa/2009/f-cp...90601e1.en.pdf

Two questions I could not find the answer to in this thread (bear with me if I overlooked them):

1. auto-trim moved the THS to up 13° while the PF was pulling up -- was the auto-trim movement a consequence of the PF's pulling up?

2. would a full down sidestick have had the authority to overcome the full up THS later on?

MountainBear wroteI beg to differ:
a. Sully landed on the Hudson with his USAF training and experience background dominating, not airline training.
b. The way for cost effective and inexpensive training starts with proper pilot selection, syllabus and training. My son following a few weeks in class then simulator sessions along one week, got his B737NG type rating and is now flying with that top notch airline. Contributors are, again: proper candidate selection, syllabus and training. Decision makers need to understand flight, airline management, and wrap-up this prohibitive cost bull.

Regarding who to spend money on, the study of ethics teaches us that the best way to get life quality is to pursue smart egotism, in which people benefit by donating private resources to their community.

Hi,

Those questions were already answered
It's YES for the 2 questions if the airplane was acting like it must be in alternate law
Pull up and the THS will go up
Pull down and the THS will go down

"You have to balance Stall Warnings in extremely rare events with far more common nuisance stall warnings, in order to maintain confidence in the system"

With an AOA probe/indicators system set up to alarm CONTINUOUSLY anytime AOA reaches predetermined level before critical AOA and BEYOND (...sorry sound like Buzz Lightyear) with the addition of a weight on wheels breaker.

Can you envisage any scenario which would lead to a "Nuisance stall warning"?


PS.-- One interesting post few pages back was from a captain who made the point that sometimes having an alarm going of continuously can detract from an individuals ability to deal with a problem. I suspect there is a balance that has to be struck and I suspect / hope that the engineers, psychologists, pilots involved in cockpit design have already thought it through.

I apologise in advance if this has been covered already, but in the absence of any trustworthy data, two buttons on the FMGC would have led them to the GPS altitude and groundspeed.

Used in conjunction with the unreliable airspeed drill, I could envisage a successful outcome to their predicament with a minimum of fuss.

On an antecedent note, don't mess with the ITCZ.

As I remember, the Air Caraibes pilots used GPS altitude to confirm that they were in stable flight using the pitch + power settings

Concerning fuel on board. Please read BEA's report appendix 7 page 116/117.
http://www.bea.aero/docspa/2009/f-cp...90601e1.en.pdf
At T/O, they had :
- 900 kg EXTRA fuel over the minimum M .82 DIRECT flight plan to CDG (alternate ORY) ;
- 1.900 kg EXTRA fuel over the minimum M .81 DIRECT flight plan to CDG (alternate ORY) ;
- 2.000 kg extra fuel over the minimum M .82 "subject to RIF to CDG" ("ETF" in BEA's wording) flight plan to BOD (alternate TLS).

At any given time, if fuel becomes an issue, the crew can decide to fly at M .81 (iso M .82) or to land at LIS/BOD/NTE without too much "embarrassment".

Concerning ITCZ crossing. Please read Tim Vasquez conclusion :
Air France 447 - AFR447 - A detailed meteorological analysis - Satellite and weather data
If CB/turbulence was a factor, why did they start to deviate at 02:08 and not during the 01:50/02:08 period (when they went through an equally "bad" area according to figure 5b shown above) ?

Concerning captain behavior : he went to rest only 10 minutes before things went bad.
Who on earth would go to rest if the 160 Nm radar picture showed some "extra-ordinary" weather ahead :ugh:

Re AOA Probe balancing
 

The F4 had an AOA probe without the need for balancing. During groundcheck we checked the free movement of the probe and it stayed where you put it.
AOA Probe

AOA value was displayed on the AOA gauge. The AOA also triggered an aural AOA tone in the headset, starting with a low frequency low repetitive tone at 15 units (if i remember correctly), becoming a steady medium frequency tone when optimum AOA was reached for landig (19.2 AOA) increasing to a high repetitive high frequency tone when AOA limits where exceeded. That was our stall warning and nothing else was needed.
Advantage over the betty bitch thing is IMHO, that you are aware over the trend the AOA is developing, whether it is increasing or decreasing.

I´m not saying that this would be the ideal system, just to make a point that the balancing problem should not lead to shutting off any kind of stallwarning below a predetermined speed. The technology is available since 1960!

Further to jcjeart's response (#1302) can someone tell me why trim is automatically triggered by nose-up or nose-down side stick commands in an Airbus? When flying my C152 I might wind in a bit of trim to ease the control loads when establishing a lengthy climb or descent but it's my decision. Once I've trimmed the aircraft for level flight I usually leave it alone regardless of altitude changes.

Hi,

Cause it's Airbus .. not Cessna ...
Can't more explain than this PDF

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

And this:
http://img402.imageshack.us/img402/8669/thsinfo.jpg

Many thanks. It will take me a while to read and assimilate :) Best wishes...

A couple of questions from an SLF
 

Apologies if these are dumb questions or have been answered before:

The BEA statement is silent on several points, presumably in order not to apportion any blame before publication of the main report in due course. One such omission is discussion of the crew's awareness of the approaching bad weather, and whether to change course. They did know they were heading towards an an area of turbulence before the captain went on his break at 01.55, but they did not decide to deviate. Would the two co-pilots have any indication after that that the storm was particularly severe? Would they have the authority to change course without consulting the captain? Would they be reluctant (more than, say a BA or Lufthansa crew) to wake him and ask to change course? Would these discussions be answered by the CVR?

THe Phantoms post a few pages back is a great summary. It sounds like a toxic mix of features that all came together in the worst possible way. The accident seems highly multi faceted, with a few major junctures to intervene ,all of which were missed. The thread has been immensely interesting. One of the most important investigations in aviation history.

Threats and errors not identified early ( aircraft design issues, flight execution) - altitude, weight, non deviation around significant dangerous weather , pitot failure, over complexity of automation and warnings ( including variation in flight Law,thrust automation, sensor and warning cut - ins and cut offs), likely significant deficiency in basic instrument scan/priorisation in the early phases of the emergency, spatial disorientation ( somatogravic- related to deceleration cues interpreted as nose down attitude), probable combined crew attention fixation AND distraction, cockpit design flaws ( sidesticks not visible to PNF, non linked controls) ,the least experienced PF the worst possible situation, non recognition of stall, seemingly simple problems with basic solutions interpreted in complex ways

Lots of little things snowballing . I only hope the final report pays full attention to every single one of the features of the accident - there appear to be so many contributing aspects.

Quick question - seeing as Alpha seems to be one of the most important metrics of flight, where an instrument is available is it not possible for it to become the measure for things like approach speeds? I.e. consider Vref crosscheck etc but set approach against desired AOA number. After all this will be consistent across weights etc. Seems it would greatly help awareness and pilot confidence in what the wings are doing.

If I may supplement what Retired F4 pointed out in re AoA indications in my response to Graybeard.
Graybeard
I don't think I am that much more knowlegable, but I am familiar with them on some aircraft.
Not really. They've been arguing that clipping the stall warning in flight at 60 knots makes no sense. I agree.

IIRC, AoA is displayed in units, but that may be aircraft specific. For the Navy Trainers I am familiar with, T-45 stalls somewhere between 29 and 30 units, the T-34 at about 26 units AoA. See also Retired F4 comments above.
OK
Roger
To the point of stall, and beyond, as your barnyard theme suggests: let it weathervane in the airflow as it will.
Yes about noise and distraction, but I see no reason to have a stall warning on the ground. There is an arbitrary decision to pick an airspeed. There is no reason to curtail it.

When you are flying, stall is about the most important thing to know about and unscrew first, since if you are in a stall, you aren't flying so much as falling.

Of all warnings, that ought to have primacy, don't you think?

Stall warning is not just another damn light or noise.
I don't agree with your assumption there. Airspeed near zero has little evidence to support it, and would have required IMO a much higher nose attitude than what FDR indicates.

I see no reason to clip AoA input based on any airspeed. That was my point. The WoW (squat?) switch would take care of the issue of spurious warnings, would it not?

AoA tied to stall warning would thus be tied to aircraft in flight/off the ground, which is where you both want and need stall warnings. It remains independent of the airspeed indicating system, as it should.

@bearfoil: I do not understand what you said regarding artificial horizon not being optioned by AF a couple of pages ago. There is one for each Pilot's Flying display, and a back up in the ISIS suite of instrumentation. :confused:

Can you elaborate? :confused:

GerardC.

We do not yet know what they saw on their radar display; that portion of the CVR where there may have been crew discussion of what was depicted on the display has not been publicly released.

We do know that a LH 744 preceding AF447 deviated by 10 NM, that an IB 340 at FL 370 and following by 12 minutes deviated by 30 NM, that an AF 330 following by 37 minutes initially deviated by 20 NM to the left of the track, and then 70 to 80 NM to the right of the track, and climbed from 350 to 370.

Vasquez estimates that the top of the Cb AF447 flew into was 56,000 feet. If a Cb of that size had been visible by daylight or moonlight, every pilot flying would avoid it.
____________________

I agree that possible fuel consumption resulting from a possible deviation was not a factor.

In the A320 Perpignan crash it seems that the sidestick couldn't overcome the fully deflected THS. How did you find out that in this case the pilots still had pitch control authority?

The report mentioned pitch down orders and the THS didn't seem to have moved.

Any clue about the THS position on the Afriqiyah A330 crashed in Tripoli?

But even after the auto-pilot/auto-throttles dropped off it took several calls to the captain to get him back in the flight station? So how bad could the turbulence have been?

Was the radar display on bright or was it left at full dim by mistake, hence maybe the 160/320 nm pic didnt show any Wx returns.These questions are difficult to answer.
When i was a senior F/O , I thought that i knew the radar like the back of my hand, only to relearn how to use it on my own as a captain.
There is [ in my opinion] some hesitation in deviating for Wx in many F/Os that i have worked with, Does not mean it applies to all.

THS position and flight law
 

jcjeant wrote
The BEA report tell us that the THS remained at 13% nose up including during the nose down stick input that occured during the descent, before 10000 ft. Therefore one (at least) of the following must hold true:
- THS movement under alternate law is more complex than suggested above; e.g. delayed, and the nose down command was shorter;
- the THS moved slightly but the BEA report does not mention it;
- flight law was not alternate law; direct? abnormal? sub-variant of alternate?
- there was a trim malfunction (but it seems unlikely that the BEA report does not mention it at this stage, so I rule this one out)

Any authoritative source on exactly how the autotrim works in the various flight laws?