Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

AF 447 Thread No. 12

Tech Log The very best in practical technical discussion on the web

AF 447 Thread No. 12

Old 31st Jul 2017, 19:03
  #1561 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
"...For every pilot, it should be bread and butter, that stall warning requires relaxing of back pressure, requires forward stick movement, not aft stick movement..."

The first Stall Warn interrupted what was a brisk movement forward of the Stick. According to the graph, the Warning stopped, Bonin continued forward stick movement, and then gently added aft pressure to stick. Don't forget, he may have suspected coffin corner, and was trying to be judicious with Pitch. He also was trying to arrest the descent his ND started, with aft stick.

Pull back, Stall. Push forward, Overspeed.

"...Perhaps you are under they impression that, while having the airplane act as a piano, falling through the air, that the crew was subjected to zero (or perhaps a little) negative g. Not so, it was falling steady state, more or less, and everything inside would be pulled into their seats by gravity with the normal load of approximately 1 g..."

Now you are in Deep Stall. With your experience of Deep Stall, and a two hundred ton aircraft squatting on a cushion of thin air, you say forward stick produces gentle response?

I would say no. A brick, or any portion of it, losing its cushion, will accelerate quickly, and perhaps even start a tumble. If 447, balanced thusly, was stable or thereabouts, do you think pilot would risk stability to gain unwanted airspeed, exacerbating the already novel and robust instability in the Roll?

But this is Deep Stall. I refer to first Stall Warn, as shown in the report.

I have flown this maneuver, albeit in a small plane. Unless rudder and aileron are strictly controlled, the aircraft acts like a man trying to stand on a bowling ball. It leaks to the side, falls forward, or even seems to slip backward. It is exhilarating, and without prior experience, I suggest would be intimidating, to say the least.

Last edited by Concours77; 31st Jul 2017 at 19:17.
Concours77 is offline  
Old 31st Jul 2017, 19:27
  #1562 (permalink)  
 
Join Date: Aug 2007
Location: The Netherlands
Age: 64
Posts: 288
Fun

Give me the airliner that produces overspeed with the nose 15 degrees high above the horizon at FL380.

Must be a mighty powerful beast to play with.
EMIT is offline  
Old 31st Jul 2017, 19:29
  #1563 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
Addendum. Recovery from this maneuver, even in a light plane, takes a lot of altitude, and gentle pitch, to avoid a second experience. That I have not done.
Concours77 is offline  
Old 31st Jul 2017, 19:30
  #1564 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
Originally Posted by EMIT View Post
Give me the airliner that produces overspeed with the nose 15 degrees high above the horizon at FL380.

Must be a mighty powerful beast to play with.
Exactly, but who knew?
Concours77 is offline  
Old 31st Jul 2017, 19:58
  #1565 (permalink)  
 
Join Date: Dec 2010
Location: Middle America
Age: 81
Posts: 1,163
But this is Deep Stall. I refer to first Stall Warn, as shown in the report.
AF447 did not enter a deep stall at 38,000 ft. It was a conventional stall. Deep stalls occur mainly on aircraft with rear mounted engines and a T-tail design. Aircraft that employ the canard design feature are prone to deep stalls if preventive measures are not included in the design. A-330s don't have a T-tail, rear mounted engines or canards and don't deep stall.
Turbine D is offline  
Old 31st Jul 2017, 20:17
  #1566 (permalink)  
 
Join Date: Sep 2016
Location: USA
Posts: 673
Yes it did, AOA for most of the ride down was about 40 degrees. Typical stall AOA is about 15 to 20.
Vessbot is offline  
Old 31st Jul 2017, 21:34
  #1567 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
Turbine D,

Point taken, but there is no other way to describe the Stall. It didn't help having full Nose Up Trim, and under slung engines at full chat (which induces nose up). Once the wing unloaded, and THS remained all NU, the elevators became trim tabs. Also, this aircraft is longitudinally stable, it resists emphatic changes in Pitch something the "trim tabs" though recorded, had little effect...upon.

Once the aircraft "settled in", it may actually have seemed docile. But the recovery shouldn't have been necessary anyway. As Wiedehopf points out, in Alternate Law, the 330 becomes a different "beast".

Last edited by Concours77; 31st Jul 2017 at 21:45.
Concours77 is offline  
Old 31st Jul 2017, 21:55
  #1568 (permalink)  
 
Join Date: Dec 2010
Location: Middle America
Age: 81
Posts: 1,163
Vessbot,
Perhaps this will help you on stall understanding:

http://www.sftentx.com/files/75234188.pdf
Turbine D is offline  
Old 1st Aug 2017, 12:47
  #1569 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by Vessbot View Post
Yes it did, AOA for most of the ride down was about 40 degrees. Typical stall AOA is about 15 to 20.
AF447 did not enter a deep stall, but a fully-developed stall.
Nothing indicates it was impossible to get out of that stall situation with proper flight control inputs.
The main culprit and obstacle was that unbelievable fully nose up auto trimmed THS ...
CONF iture is offline  
Old 1st Aug 2017, 14:25
  #1570 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
Deep Stall. Nomenclature. Critical to a technical approach, without doubt.

1. How did it get there?
2. What made it (functionally) not recoverable?
3. What issues in the AC design made it (inadvertently) fatal?

Three poorly designed, inadequate pitot systems, due to be replaced by a line that did not take seriously the vulnerability of the aircraft in conditions that were frankly typical of the airspace. Threat of a pilot strike finally got them to take action....

Full thrust, bleeding energy, loss of lift at the outer wing, Chronic NU command. Direct Law in Roll, the betrayal of the STALLWARN system, turbulence, cockpit disarray (visitor). Delayed recognition of the loss of speeds display and Normal Law by pilot crew. (Fifteen seconds). One can make the argument that the loss of speeds and Normal Law, caused by Thales Pitots, was the direct and sole cause of the ultimate impact with the ocean.... "Shoulda, Woulda, Coulda...."

Then the piece de resistance, a full up HS.
Stall entry unnoticed, and a mental lock on issues other than those that would cause a recovery.

The word "cascade" comes to mind. This is not a simplistic and isolated wreck.
Concours77 is offline  
Old 1st Aug 2017, 15:03
  #1571 (permalink)  
 
Join Date: Dec 2006
Location: Florida and wherever my laptop is
Posts: 1,350
It looks very much like flying 'on the wrong side of the drag curve' this is a stable state and if your aircraft has sufficient power then the lift is supplemented by the power from the engines as they have a significant upward vector due to the nose high angle and the aircraft will happily fly straight and level at below its normal stall speed 'sitting' on the power. Can be fun if you can see what you are doing.
If the aircraft is at height and power to weight ratio is not good or power is reduced the aircraft will drop - quite fast - as AF447 did. The nose high attitude must be maintained and for that the tail plane had motored into full nose up 'trim' - and every throttle increase with underslung engines also assisted the nose up attitude. This is not a classic stall.

The aircraft is now in a stable nose high state dropping faster than its forward speed. Unwittingly, PNF by insisting that PF kept the wings level assisted in the maintenance of the stable state. Had PF lost the roll control the aircraft would have rolled out of its stable state - perhaps before the tail had motored to full NU. That would have provided a more 'conventional' LOC that they may have recovered from. As it was they sat on the wrong side of the drag curve all the way to the surface.
Ian W is offline  
Old 1st Aug 2017, 15:20
  #1572 (permalink)  
 
Join Date: Jan 2008
Location: uk
Posts: 860
Originally Posted by Vessbot View Post
Bombardier modified the C* (C-star) FBW design control system to be speed-stable,
This is what's also in use by the 777 and 787. C*U
Before even hinting that C*U might be a solution to AF447, one should consider what it becomes when U (ie. speed) is unknown...

787 is, I believe, slightly different in that if speed is unavailable it calculates a synthetic airspeed from AOA and other data both for display and use in the control laws. This might have been developed with an eye on the vulnerability of C*U to loss of U.

C*U hasn't been without its problems though:
https://www.flightglobal.com/news/ar...d-data-423735/
https://www.atsb.gov.au/media/577302...-149-final.pdf

Pitot icing, plane's computers can't cope, yet again (even with the synthetic airspeed).

to quote the second link:
In response to a previous, similar event on another B787-8, the FAA published an airworthiness directive warning flight crew not to make large abrupt magnitude flight control inputs in response to
unrealistic drops in airspeed.

Boeing also revised the flight control software to limit the rate of elevator feel reduction with drops in airspeed. This will specifically allow the column to stay at a higher feel force to mitigate large and abrupt unintentional pitch inputs.
So, C*U is not quite the fix-all some seem to think it is.

Another salient quote from that report:
In this case, the crew showed a high level of professionalism in response to a weather related event. The crew demonstrated high levels of communication and coordination, promptly applied checklists and procedures.
And therein, I submit, lies the crucial difference between making it home and ending up at the bottom of the ocean.
infrequentflyer789 is offline  
Old 1st Aug 2017, 18:12
  #1573 (permalink)  
 
Join Date: Sep 2016
Location: USA
Posts: 673
You guys seem really wedded to the definition of "deep stall" being limited to the particular blanked stabilizer phenomenon in T-tails. Or otherwise diminished pitch authority. Fine, I'll live with it. Nevertheless, this airplane was very deeply stalled.

And Turbine D, there's no need to act like an about it.

Originally Posted by Concours77 View Post
I have flown this maneuver, albeit in a small plane. Unless rudder and aileron are strictly controlled, the aircraft acts like a man trying to stand on a bowling ball. It leaks to the side, falls forward, or even seems to slip backward. It is exhilarating,

[...]

Also, this aircraft is longitudinally stable, it resists emphatic changes in Pitch something the "trim tabs" though recorded, had little effect...upon.

Once the aircraft "settled in", it may actually have seemed docile.
This is an interesting point. Normally, if you asked anybody, they would say that a transport airplane should have docile stall characteristics, so as to limit workload (and potential for further upset) in a recovery. Now, it seems to me at least plausible that this docile behavior contributed to a lack of stall identification by the crew. Talk about a double edged sword!

Originally Posted by Concours77 View Post
Perhaps better to enlist PPLs to fly this family? At least the Stall would be no less baffling to them. Perhaps retire them out of type after 2000 hours?
Well, everything I wrote about the airline flying experience wrt. the maintenance of basic flying skills outside the normal airline profiles, applies equally to non-Airbus types. So, we'd be in a real pickle if we apply that solution industry-wide! (I know you're being tongue in cheek, of course.)
Vessbot is offline  
Old 2nd Aug 2017, 14:56
  #1574 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
"...Because the Aircraft remains safe even when the pilot pulls on the stick to the stop thanks to the high angle of attack protection, Airbus position is that no additional alerts about low speed/energy situations is needed at high altitude because large altitude loss due to stall cannot occur on Airbus Fly-By-Wire aircraft.

Moreover, no additional stall warning or stick shaker is needed with this kind of protection. Stall warning is only re-introduced on airbus aircraft when the angle of attack protection is lost (loss of normal laws)...."


ASHWG LAA Report_15 March 2010 Final Version_Appendix 1

Fair enough. What of performance after reversion to alternate Law? Stall Warn should compel action. If ignored, or dismissed, there is no other resolution for Airbus pilot. Shaker is unmistakable, irrevocable and emphatic compulsion to act.

So. Why the waiver from FAA? I would understand the waiver if clearly limited to Normal Law, the alpha Prot is its own Stall Warn, and cannot be overridden.

Alternate Law is a "different" airplane, should it not be required to meet minimum standard of other aircraft?

Roll in Direct? Perhaps a clever way to introduce instability in Roll, and a clever way to provide Nose Down, independent of crew? Unstable roll axis is perhaps the only Stall cue they would receive? Did Bonin outfox the airplane, and himself, by learning to manage the new roll? Or was the aircraft supposed to roll until the nose dropped, by design? Stick and rudder by program?

What this crew needed was instability in the descent, not docility. They were focused on stability as always, like all good Airbus drivers?
Concours77 is offline  
Old 5th Aug 2017, 18:55
  #1575 (permalink)  
 
Join Date: May 2006
Location: Genf
Age: 55
Posts: 13
I ve tested several times a very good malfunction in the box: Volcanic ash. It is a nasty one because both engines are not usable and all your pitots are blocked. 777 is a democracy, if 2 out of the 3 pitots are blocked the ADM that are in majority reporting the same speeds are "winners into the election" for best data. In other works you will be prompted with erroneous speed for a while, till you hit 80 kts IAS and then the last ADM will be "elected". Back to that malfunction, your best friend would be the weather briefing with winds (predicted) and GPS ground speed since you cannot apply pitch, power, performance trick.
Houba is offline  
Old 6th Aug 2017, 00:32
  #1576 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
Houba

IRU? Why waste inertial data. Ice and ash immune...
Concours77 is offline  
Old 8th Aug 2017, 21:40
  #1577 (permalink)  
Thread Starter
 
Join Date: Oct 2013
Location: Wengen
Age: 51
Posts: 377
WX

Preflight WX: charts were outdated and crew simply could not read them properly as on page 5.
AF are in the stone age regarding WX as they did not do provide any satellite indications!
AF 447 LE PLAN DE VOL AF447-Plan de vol. - ppt télécharger
Winnerhofer is offline  
Old 9th Aug 2017, 22:57
  #1578 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 78
Posts: 1,412
re-hash of old stuff

Great point(s) Vessbot

You guys seem really wedded to the definition of "deep stall" being limited to the particular blanked stabilizer phenomenon in T-tails. Or otherwise diminished pitch authority. Fine, I'll live with it. Nevertheless, this airplane was very deeply stalled.

And Turbine D, there's no need to act like an about it.
Years ago, like thread 5 or 7 or 9, I tried to make the point that the 'bus must have nice stability and low buffet when at a high AoA. So defeating any of the remaining "protections" ( God, how I hate what that term implies) by holding a nose up command as the speed rapidly decreases makes for a "smooth" stall entry and except for the degraded roll control laws the sucker just sits there happy and descending at 10,000 feet per minute.

I can understand that the crew could not readily accept that they were deeply stalled. Ask a "delta" pilot like me in my early years, of any type - F-102, Mirage, F-106, Vulcan, Concorde, and so on. Very smooth transiton from lift to no lift and plenty drag and vertical velocity you do not want. My recollections of the Deuce was it had a "buzz", but no wing rock or buffet that I had in the other 3 or 4 planes I flew. But the bottom fell out on short final if you pulled back a lot with low power. So I cut some slack for the crew recognizing their stall entry, but only a little. Lack of understanding the confusing degraded control laws and failure to just hold whatever control inputs were there when the aero sensors went tits up and the AP kicked out did not help at all. Lastly, as CONF pointed out, the THS trimming while in a backup mode was the icing on the cake.

I am not gonna demand all the newer folks here go back thru the thousands of posts where we dissected this terrible crash, but maybe using the "search" features one could see "deep stall", "longitudinal stability", control law regression, and many more aspects of the crash.
gums is offline  
Old 9th Aug 2017, 23:52
  #1579 (permalink)  
 
Join Date: Nov 2016
Location: Lakeside
Posts: 428
Moreover, no additional stall warning or stick shaker is needed with this kind of protection. Stall warning is only re-introduced on airbus aircraft when the angle of attack protection is lost (loss of normal laws)...."

ASHWG LAA Report_15 March 2010 Final Version_Appendix 1

Fair enough. What of performance after reversion to alternate Law? Stall Warn should compel action. If ignored, or dismissed, there is no other resolution for Airbus pilot. Shaker is unmistakable, irrevocable and emphatic compulsion to act.

So. Why the waiver from FAA? I would understand the waiver if clearly limited to Normal Law, the alpha Prot is its own Stall Warn, and cannot be overridden.

Alternate Law is a "different" airplane, should it not be required to meet minimum standard of other aircraft?

Does there not seem to be a large omission here, gums?

Each time the crew experienced the STALLWARN, it should have been preceded by a stick shaker.

If, in AltLaw, the "cannot Stall" morphs into an aircraft that can not only Stall, but does so so subtly that a separate device must be required ?

It occurs to me that if the shaker was shaking without stop all the way down they may have tumbled to the state of the aircraft?
Concours77 is offline  
Old 10th Aug 2017, 09:22
  #1580 (permalink)  
 
Join Date: Jun 2011
Location: West of Offa's dyke
Age: 85
Posts: 476
There appears to be some misunderstanding of the character of high altitude/high Mach No stalls on current transport aircraft. On designs using supercritical wing sections on which upper surface flow breakdown starts at or near the TE and for which the design span loading leads to the critical maximum local lift coefficients being located about two thirds semispan the loss of lift is only slightly aft of the CG, so any pitch up is indeed mild. This is true no matter which side of the Atlantic the design comes from. The limiting factor in cruise conditions is buffet and in fact “stall” is usually defined by buffet reaching +/- 1g vertical acceleration at the pilot’s seat. The A330 high altitude/cruise Mach No stall data published in the second “Expert’s Report” (in French only unfortunately) shows this occurred at about 9 deg AOA on a simulated AF447 test, accompanied by +/- 0.5g lateral acceleration. I’m no pilot, but surely this sort of environment is not “benign”?. The authorities obviously think that this level of buffet plus a verbal stall warning announcement constitutes compliance with FAR 25.201 d (2).
The recent publication of a joint Boeing/Airbus paper “Stalling Transport Aircraft” (Society of Flight Test Engineers) makes interesting reading in this respect. For those who don’t want to bother, here are some relevant extracts (my emphasis)
<At high altitude/high Mach number, the Vs1g cannot be determined due to early triggering of buffeting. This buffeting is caused by local shock waves which excite the structural modes of the aircraft. Theoretically a ClMAX may exist at high Mach number but the associated level of buffeting prevents efforts to identify it during flight tests for aircraft structural integrity and safety reasons.
Nevertheless, according to the regulation a minimum of 1.3g maneuverability up to “buffet onset” must be demonstrated for each flyable Mach number. Therefore the ClMAX for a given Mach number will generally be defined by the level of buffeting corresponding to ±0.1g vertical acceleration level measured at the pilot’s seat (so called “buffet onset”) and is named ClBUFFET

Pitch up
“Pitch-up” can be observed during stall mainly on aircraft fitted with swept wings. It is due to the sudden loss of lift on the outer part of the wings, which creates a nose-up moment (see Figure 7). This phenomenon can also occur at high altitude/high Mach well before ClBUFFET is reached and is again due to shock waves destroying the lift on the outer part of the wings.

Expected 1g stall characteristics of large aircraft
Typical stall characteristics of transport aircraft in 1g non-accelerated flight (ref FAR Part 23.201) begin with the onset of initial buffet. This is best described as light airframe buffet which begins a few knots prior to stick shaker. As the aircraft approaches Clmax the level of buffet generally increases and can become severe to deterrent in nature. It is not uncommon to see buffet with a repetitive load factor of ± 1 G in the vertical direction and ± .5 G in the lateral direction (see Figure 20). It feels similar to driving an automobile across railroad ties. Buffet on large airplanes tends to be much greater than experienced in smaller aircraft. This is due to wing airflow separation and turbulent airflow vortices which produce a strong excitation forcing function on the wing. This excites the fundamental frequency of the fuselage leading to large vertical and horizontal deflections. It can be very evident on the flight deck, where anything not securely tied down, such as an errant water bottle, can get hurtled into the air.
Stall identification is deterrent buffet for most recent models in the clean wing configuration
Certification Stall Requirements for FBW Aircraft
Stalls for certification are performed initially at forward CG to determine the stall speeds that will be used for speed computations. For FBW aircraft, stalls must be performed in Normal Law to demonstrate the ability to control the aircraft throughout the stall and recovery, should an inadvertent stall occur despite the protections (for example, in the case of severe windshear). Because the angle of attack limits in Normal Law prevent the aircraft from stalling, a specific version of the Normal Law is created which shifts the limiting AOA to a higher value (usually by 10 degrees). Additional stalls are then performed in Degraded Control Laws, in all configurations, and at aft CG. For stalls in Augmented Control Laws, it is essential to place the stick forward of neutral during recovery, since the C* law is g-demand law, and neutral stick is a 1.0g command. If the stick is not placed forward of neutral, more nose up elevator than desired will be applied at the stall break.
Stalls at high altitude are not required in the classic sense. Instead, a series of wind-up turns are performed at constant Mach number to determine buffet boundaries. This determines how much g the airplane can sustain before the airplane begins to buffet, which is useful to the operational pilot when flying at maximum altitude for a given weight. If the airplane is flown too slowly, there will be an increase in buffet to the point that altitude will have to be sacrificed to regain speed. The same holds true if the airplane enters turbulence which applies g loads higher than the buffet boundary for a given Mach/altitude/weight combination
Stalls at High Mach Number
Stalls at Mach numbers normally associated with cruise flight (0.78-0.89 Mach) are not possible in level flight because Mach number decreases as the airplane decelerates. This increases Clmax and effectively increases the margin to the stall. It is difficult to tell when the test point will end, because the end point is shifting. Angle of attack limits for safety are equally difficult to predict. As the airplane decelerates, the level of buffet increases significantly and rapidly becomes deterrent buffet. The g-break may be difficult to recognize, either from the g trace, or the Cl trace, therefore a rapid change in vertical velocity may be the first good cue of the stall. Pitch-up may be present during deceleration, complicating the pilot’s ability to smoothly control pitch with increasing levels of buffet.>

Last edited by Owain Glyndwr; 10th Aug 2017 at 09:38.
Owain Glyndwr is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information -

Copyright © 2021 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.