PPRuNe Forums

PPRuNe Forums (https://www.pprune.org/)
-   Tech Log (https://www.pprune.org/tech-log-15/)
-   -   Habsheim (https://www.pprune.org/tech-log/528034-habsheim.html)

CONF iture 16th Dec 2013 01:50


Originally Posted by Chris Scott
Note that this BEA description does not specify that an alpha of 17.5 deg will be achieved if the pilot maintains full back-stick.

The BEA does not elaborate at all but applying full back stick is what it takes to go to alpha max.

Some readers may be surprised to learn that at t = 0 (defined as impact) the stick was still not fully back. IIRC, full deflection is 20 degrees. The last recorded stick position was -17 deg (17 deg nose-up command)
Full deflection in roll is 20 deg but 16 in pitch. The stick was fully back.

balsa model 16th Dec 2013 03:00

So to get to alpha max, one has to apply sustained full back stick. I'm guessing that this means that as the full back stick is applied, the alpha limitter progressively increases its limit value from alpha prot to alpha max.
Out of curiosity, how much delay would this entail?

Also, I recall something about Airbus fbw going to "direct law" at expected landing radar altitudes, and that this mode gives the pilot direct stick-to-pitch control. Could that have been the pilot's motivation for going so low? To have a better feel over the remaining pitch authority?

CONF iture 16th Dec 2013 03:24


Originally Posted by balsa model
Would the extra couple of degrees of AoA, that are presumed to be held back by the s/w (is that you point, CONF iture?), have changed anything for the better in the outcome?

To be clear, I'm questioning why BEA+Airbus kept absolutely silent on the fact that the airplane had no intention to deliver anything more than alpha prot despite the pilot request.
But to answer your question, my opinion is that it would have made the difference considering that according to the report the engines have been registered at 91% N1.

IF the extra increment in lift extended the flight long enough for the engines to spool up, wouldn't the go-around thrust-pitch couple just push the airplane into a stall? The crash would then be tail-first, perhaps.
No, alpha max is part of the protection, and so whatever the thrust setting.

vilas 16th Dec 2013 04:47

Dozy, Chris r r r
I think unnecessary lengthy technical discussion is going on when there are some simple aviation facts staring in the face. I do not know what was planned but you can see what was executed. The pilot decided to slow the aircraft to minimum speed taking advantage of the AOA protection, fly level and go around. In any aircraft primary instrument of flight path change is pitch. The thrust maintains speed. The pitch up caused by thrust is secondary effect. Incidentally it is dampened in A320 by FBW and you need to pitch up. Not doing so was a possible cause of another crash. When pilot was late in setting the precise thrust to fly level at 100ft the aircraft sank. He could not pull back because he was at MAX or near MAX AOA so unable to change the flight path. Thrust was the only alternative but being at idle for a while it takes about 4 seconds to get out of idle and when you are sinking from 40ft, 4seconds must have seemed like a life time. So he panicked and thought ATHR was not behaving and moved the thrust levers manually. Alpha floor should have come between alpha prot and alpha max. Now the question is how and when exactly was it disconnected? There are three possibilities.
1. It was disconnected permanently before slowing down started. This does not seem probable as the pilot expected thrust activation.
2. It was disconnected as it got activated between alpha prot and alpha Max. In this case alpha floor will remain dormant till the speed increases beyond Valpha prot.
It is possible pilot did not know this considering the technology was new and that is why he was surprised by lack of response.
3. He may have been holding on to the disconnect button to prevent alpha floor till speed came to alpha max and inadvertently exceeded 15 seconds because he was preoccupied visually and disabled it.
In any case he needed quick response to gain those critical extra 50ft which was not possible because of the max angle of attack.
Why should a line pilot who only has access to FCOMS indulge in manoeuvres which even the test pilot did not, in radically different aircraft? I am unable to comprehend that.









Clandestino 16th Dec 2013 06:39

ATHR disabling (by holding disconnect buttons for more than I-forgot-the-exact-number seconds) was planned but not executed. Alpha floor didn't activate because it's inhibited below 100 ft per design, in order not to spoil the landing.


Why should a line pilot who only has access to FCOMS indulge in manoeuvres which even the test pilot did not, in radically different aircraft?
Showing off. Careless planing combined with sloppy execution. Chances are they never meant to fly that low, which does not imply it is acceptable to make high alpha pass with planeload of passengers at 200 instead of 20 ft.


Also, I recall something about Airbus fbw going to "direct law" at expected landing radar altitudes, and that this mode gives the pilot direct stick-to-pitch control.
Some widebodies stop autotrim, narrowbodies add automatic pitch-down command to facilitate the flare but going to direct law happens only with gear extension when aeroplane is already in alternate law and that was definitively not factor here. Aeroplane was totally sound just before controlled flight into trees.

rudderrudderrat 16th Dec 2013 09:35

vilas

Alpha floor should have come between alpha prot and alpha max. Now the question is how and when exactly was it disconnected?
We don't need to discuss that - because as Clandestino points out, "Alpha Floor is available, when the flight controls are in Normal Law, from lift off to 100 ft RA at landing." (FCTM)

Alpha Floor was never going to activate whilst they were below 100 RA.

vilas 16th Dec 2013 11:26

r r r
If the speed went below alpha prot only when they went below 100ft then offcourse alpha floor was disabled. Did the pilot know that? The Airbus design is such that it is possible to fly hundreds of normal routine operations without much knowledge. I consider that as a positive but the pilots have to understand that if they try out something beyond normal they could be asking for trouble because that requires much deeper study of the design and its implications which is beyond the scope of FCOMs and line pilots. Protections on civil transport aircraft are not built to routinely fly the aircraft on the brink but to save the situation when forced into it by abnormal factors. At Perpignan they planned it then abandoned it and last moment executed it at low level and were not able to handle the consequences.

rudderrudderrat 16th Dec 2013 11:59

vilas

If the speed went below alpha prot only when they went below 100ft
From page 40 of the BEA report, at time 1245
100 RA Speed 151 kts.
48 RA Speed 141 kts.
40 RA Speed 137 kts.
30 RA Speed 116 kts.

It appears to me that the aircraft was deliberately flown below 50 RA with the speed allowed to wash off, apparently waiting for Alpha Floor to "save the day".

then offcourse alpha floor was disabled. Did the pilot know that?
Probably he did not know that, as the FCTM was published later, and I still can't find a reference in FCOM.

the design and its implications which is beyond the scope of FCOMs and line pilots.
I agree.

Chris Scott 16th Dec 2013 12:13

Given the briefed game-plan...
 
vilas,

As you know, this thread was started from the transfer of a few posts on the AF447 Thread No. 11. (Not my idea!)

If you look again at post #1, you may be reminded that - in an earlier post on the AF 447 thread - I had opined that the crew had been relying on Alpha-Floor to initiate the G/A (i.e., when the AoA reached +15), but that they had allowed the a/c to sink below the Alpha-Floor inhibition height without realising the consequences. That assumption on my part was naive, and on reflection would have made it very difficult for the captain to go around at a position (in relation to the crowd) of his choosing. Alpha-floor (+15 deg) is only half a degree above alpha-prot.

As CONF_iture pointed out, however, the captain had specifically briefed that he would be inhibiting Alpha-Floor, in order to continue pitching up to alpha-max. His plan was to maintain alpha-max (therefore necessitating full back-stick) in level flight (he said at 100ft), using manual thrust to maintain height. (Note that, at a stedy AoA, thrust controls VS, not speed - rather more like a light piston-engined a/c on the approach.)

The two key defficiencies in the execution of the briefed game-plan seem to have been:

(1) Rushed, straight-in approach, arriving over the airfield boundary (inbound) with too much energy to allow a reduction in speed to the planned Valpha-max during the transit of an unfamiliarly-small airfield. This may have tempted the PF to delay the go-around while he waited for the IAS to decay to an acceptable (in his terms) value.

(2) Descent below 100R. From the time they were approaching 100 ft on the approach until about t -3, the PF consistently did not pull hard enough on the stick to maintain or recover to the briefed height. This apparent reluctance to pull harder is all the harder to explain, because the PF was looking for the highest-possible deck angle to show the crowd.

Quote from CONF_iture:
Originally Posted by Chris Scott:
"Note that this BEA description does not specify that an alpha of 17.5 deg will be achieved if the pilot maintains full back-stick."
The BEA does not elaborate at all but applying full back stick is what it takes to go to alpha max.

Yes, we are in accord on that one. :ok: However, in the earlier post you've quoted, I was just trying to point out that the BEA seems to have carefully emphasised that alpha-max (in this case 17.5 deg) will not be exceeded, without stating that it will necessarily be precisely achieved.

Quote from CONF_iture:
"Full deflection in roll is 20 deg but 16 in pitch."

Thanks, I haven't found that figure yet. The DFDR trace seems to shows a rapid rise to about -17, at which point the trace ends without becoming steady.

Quote from CONF_iture:
"The stick was fully back."

As I've already said:
"Between t -9 and t -4, the stick command was between -6 deg and -7 deg, but this was relaxed slightly at t -4, falling to -5 deg.
"The last recorded stick position was -17 deg (17 deg nose-up command), rising rapidly from -5 deg at about t -3."

The back-stick reached about 17 deg (presumably maximum), but not for a measurable period prior to impact. That's the reason for my observation in (2), above. Why he did not pull harder from t -26 to t -3 is inexplicable, if only because it was delaying the desired attainment of alpha-max. The modest back-stick up to t -4 would, as you know, never achieve alpha-max.

vilas 16th Dec 2013 12:40

r r r
I am unable to paste the report in google for translation. Below is what you wanted.
DSC-27-20-10-20 P 4/6
In addition, below 200 ft, the angle of attack protection is also deactivated, when:
‐ Sidestick deflection is less than half nose-up, and
‐ Actual α is less than αprot –2 °
Note: 1. At takeoff α prot is equal to α MAX for 5 s.
2. αfloor is activated through the A/THR system, when:
‐ αis greater than αfloor (9.5 °in configuration 0; 15 °in configuration 1, 2; 14 °in
configuration 3; 13 °in configuration FULL) or
‐ Sidestick deflection is greater than 14 °nose up, with either the pitch attitude or the
angle-of-attack protection active.
The αfloor function is available from lift-off to 100 ft RA before landing.

Chris Scott 16th Dec 2013 13:22

FCOM information at time of accident
 
Quotes

From rudderrudderrat:
"It appears to me that the aircraft was deliberately flown below 50 RA with the speed allowed to wash off, apparently waiting for Alpha Floor to 'save the day'."

From vilas:
"If the speed went below alpha prot only when they went below 100ft then of course alpha floor was disabled. Did the pilot know that?"

Response from rudderrudderrat (my emphasis):
"Probably he did not know that, as the FCTM was published later, and I still can't find a reference in FCOM."

Firstly, the evidence strongly suggests that he was not relying on Alpha-Floor; contrary to what I originally thought.

Secondly, I think he should and would have known of its inhibition below 100R.

Regrettably, I have mislaid the Airbus FCOM volumes dished out to us BCAL (British Caledonian) pilots on arrival at Blagnac for our conversion course on 1/1/1988. Prior to that, we had perused the BCAL Ops Manuals (our own FCOM) dated August 1987. After finishing our ground course and simulator (we were just a week behind the first Air France course), BCAL issued a wide-ranging amendment, dated 17FEB1988, which included considerable changes to the Flight Controls chapter in the Technical Manual. The changes were partly necessitated by changes in FBW logic (Dozy please note), and partly an improvement in the explanations and diagrams.

At both 25AUG1987 and 17FEB1988, the BCAL chapter on Flying Controls describing the Alpha protections aparently fails to mention that Alpha Floor is inhibited below 100R.

However, the BCAL chapter on Power Plant - which includes a description of A/THR, and which remained unchanged from 25AUG1987 in the 17FEB1988 amendments - includes the following:

A/THR can be engaged:
- Manually, by.....
- Automatically, when the pilot initiates a T/O or G/A....
or - if there is an alpha-floor detection after lift-off and down to 100 ft (R.A.) on landing.
In the particular case of alpha-floor detection, the MAX T.O. thrust is automatically selected whatever the position of the thrust lever.

DozyWannabe 16th Dec 2013 14:26


Originally Posted by vilas (Post 8205755)
I do not have access to English version of the report.

That's because there isn't one - as I've said before - all I've had to work with over the years are extracts - but I've managed to OCR and translate some of the lesser-known aspects of the report recently.


Originally Posted by CONF iture (Post 8207013)
Not more surprising than most of you who have read or not the report are still pretending that the airplane was at alpha max.

To the best of my knowledge, no-one's claimed that, and we've already confirmed that the FCTM does not state that pulling back in High AoA Protection mode will definitely give Alpha Max - just the closest to that value that the systems can give with the flight parameters at the time. If a pilot assumes that it will give them Alpha Max precisely, then they are in error.


Originally Posted by tdracer (Post 8207118)
OK, I'm going to duck for cover after I post this, but....

There were reports that the DFDR and CVR didn't jibe - that there was a 3 second discrepancy.

Nope - that was based on a misinterpretation of the data by the retired investigator the lawyers used to give a second opinion. Ray Davis was an excellent investigator, but he had no prior experience with DFDRs.

Posted a while back:

Originally Posted by me
DFDR matches CVR. If you are referring to Ray Davis's independent reading of the FDR, he was not experienced in dealing with the new digital models, and got it wrong.

From the document posted by Franzl months ago : http://www.crashdehabsheim.net/Rapport%20Airbus.pdf


Mr. Davis was apparently not aware of the convention (which is apparently unique to France) that requires that the transcripts of forward accelerations are shown with a negative sign. [He is] therefore claiming that in the last seconds the negative acceleration shown in the transcript demonstrates that the aircraft was decelerating and therefore one or both engines were not providing sufficient thrust.

The flight recorders did not stop instantaneously at Habsheim. In the final report produced by the Commission of Inquiry it clearly states that -after the first impact with the trees, the CVR continued to operate for around 1.5 seconds and then stopped. The DFDR continued to operate
for around one second [after impact] then gave incoherent data for around two seconds". The exact cause as to why the recorders stopped almost simultaneously before the aircraft finally came to rest could not be determined. The most probable cause is that the power supply cables of the two recorders broke.

Additionally, as the (admittely poorly-)translated "Other Factors" section I posted earlier confirms, the official BEA report does not find that the factors leading to the crash lay with the flight crew alone - the sequence began with poor preparation and briefing materials on the part of the airline. The BEA's report does not explicitly apportion responsibility - because like other agencies that are nominally independent, yet still civil service (like the UK AAIB, ATSB etc.), their remit does not extend far enough to allow them to do so. Of course, the legal teams were counting on the press they were briefing to be unaware of that distinction.

Because the crew's lawyers were arranged by the union, it could be argued that they tried to paint a picture of a cover-up on AI's behalf, because to draw attention to the report's actual content - i.e. implicit but robust criticism of AF - would be biting the hand that fed the union's members.

For their part, AF's defence during the criminal and civil proceedings involved falling back on an old document that specified the airshow "hard floor" (i.e. minimum altitude) to be 600ft. This raises the question of how, if that rule was still in effect, they were able to sign off the Habsheim display in the first place - and it also demonstrates that it was in fact AF who were primarily responsible for the effort to have the crew "thrown under the bus".

As for "how did they end up so low and slow?", as I understand it the chain of events was set in motion during the (rushed) approach. It wasn't that the PIC was waiting for an "acceptable" (as in sufficiently breathtaking for the spectators) altitude/speed during the flypast, it was because they were so high and hot at the start of the final that they chopped off too much power.

CONF iture 16th Dec 2013 18:39


Originally Posted by balsa model
So to get to alpha max, one has to apply sustained full back stick. I'm guessing that this means that as the full back stick is applied, the alpha limitter progressively increases its limit value from alpha prot to alpha max.
Out of curiosity, how much delay would this entail?

Fast process - The protected aircraft takes advantage of the electronics to get the maximum performance in the minimum time and outperform the non protected aircraft.


Originally Posted by Chris Scott
Thanks, I haven't found that figure yet.

I have some unofficial references for now, but I will try to locate and post some official ones ... Wilco

The back-stick reached about 17 deg (presumably maximum), but not for a measurable period prior to impact. That's the reason for my observation in (2), above. Why he did not pull harder from t -26 to t -3 is inexplicable, if only because it was delaying the desired attainment of alpha-max. The modest back-stick up to t -4 would, as you know, never achieve alpha-max.
I think it was all about the show ... Even if the planned altitude was for 100 ft AGL and the pilot pretended he thought he was maintaining just that, I would not be immensely surprised if his real intention was to go lower ... (?)
Pulling harder early in the sequence would have made the aircraft that was too fast climb far too early to his liking.
Late in the sequence, the pilot, now fully aware of the approaching trees, but well aware that thrust was not coming as he expected it, delayed as long as possible the pull up request.

Regrettably, I have mislaid the Airbus FCOM volumes dished out to us BCAL (British Caledonian) pilots on arrival at Blagnac for our conversion course on 1/1/1988.
I certainly would not blame you for courageously making space ...

DozyWannabe 16th Dec 2013 19:26


Originally Posted by CONF iture (Post 8210064)
Fast process - The protected aircraft takes advantage of the electronics to get the maximum performance in the minimum time and outperform the non protected aircraft.

Real Time process to be more precise. Actual (as opposed to theoretical) Alpha Prot, Floor and Max boundaries are calculated constantly in real time. The values are functions - not hard-coded constants.


I have some unofficial references for now, but I will try to locate and post some official ones ...
I'd be as interested in the unofficial values (and where they came from) as I am in the official ones.


I would not be immensely surprised if his real intention was to go lower
I'd be very surprised - as the Airbus test pilots who came up with the scenario were very specific about not going under 100ft RA to maintain a reasonable safety margin.


Pulling harder early in the sequence would have made the aircraft that was too fast climb far too early to his liking.
With Alpha Floor disabled, the aircraft would maintain current altitude at calculated max AoA up to theoretical Alpha Max - it wouldn't climb, if the documentation is accurate.


Late in the sequence, the pilot, now fully aware of the approaching trees, but well aware that thrust was not coming as he expected it, delayed as long as possible the pull up request.
Can you confirm in the FDR trace?

(I'm still waiting for the source that gives you 2.5deg, by the way...)

Additionally, if the PIC was expecting thrust sooner than it was provided, then he was clearly ignorant of the properties of high-bypass turbofans. Either that or ignorant of the consequences (namely allowing the engines to spool down) of the thrust settings he ordered to make the approach.

HazelNuts39 16th Dec 2013 20:26


Originally Posted by DozyWannabe
With Alpha Floor disabled, the aircraft would maintain current altitude at calculated max AoA up to theoretical Alpha Max - it wouldn't climb, if the documentation is accurate.

It will only do that at constant airspeed, i.e thrust equal to drag. With thrust less than drag the aircraft decelerates and AoA has to increase progressively to maintain lift equal to weight, i.e. the sidestick has to move progressively backwards. When alpha-max is reached and the aircraft is still decelerating it will descend.

DozyWannabe 16th Dec 2013 20:37


Originally Posted by HazelNuts39 (Post 8210263)
It will only do that at constant airspeed, i.e thrust equal to drag. With thrust less than drag the aircraft decelerates and AoA has to increase progressively to maintain lift equal to weight, i.e. the sidestick has to move progressively backwards. When alpha-max is reached and the aircraft is still decelerating it will descend.

Correct - but in this case the aircraft had finished decelerating before it crossed the grass strip threshold. If the aircraft's forward momentum was still subject to cutting power then it would continue to decelerate, but it did not - it maintained the airspeed determined by the approach settings.

RetiredF4 16th Dec 2013 21:24

Flyover altitude
 
When the accident happened just in front of my homebase, we fast jet pilots discussed the possible traps of this failed display long hours. That short time after the accident we had no official information available except the location and environment of the nearby accident site. We could look at the crash site when departing visual from our base.

Any pilot of my outfit was familiar with low flying down to 100 feet and up to 540 KIAS in Labrador, and we all agreed that flying an airliner at 100 feet or below over a small airfield at minimum speed was suicidal planing.

We discussed, why the crew even went considerably below that planned altitude and came up with the suggestion, that it might not have been the intention of the crew, but happened due to the perspective of the small little airfield. Flying a visual approach to a small little runway / airfield like Habsheim when being used to bigger airfields and buildings like Basel just minutes before gives the visual illusion of being at a higher altitude above ground than actual. Which raises the question, why the information by the Radar Altimeter was neglected. When you expect in advance that the thing will give altitude warnings you have no intention to obey to (as the intention was to do a visual pass and not one based on inside cockpit information) the information is filtered to zero as non essential background noise by the brain.

This flyover should never have happened, we would never have gotten the permission to execute such a flyover with our jets. In my view all the speculation about possible technical aspects is noise to cover up the fact, that the crew f** ed up badly and didn't register the descent below 100 feet. Only altitude information comes from the Radar altimeter, Captain and FO did neither acknowledge nor react to those warnings.

DozyWannabe 16th Dec 2013 21:41


Originally Posted by RetiredF4 (Post 8210445)
We discussed, why the crew even went considerably below that planned altitude and came up with the suggestion, that it might not have been the intention of the crew, but happened due to the perspective of the small little airfield.

Not even that. As far as I can tell from the report, AF signed off on the profile provided that the flypast was over the paved strip (which had no obstructions in the run-off area) and the crew were comfortable with the profile. The briefing materials provided were inadequate - specifically the photocopies of the airfield charts provided to the crew effectively deleted the graphics indicating the trees.

In the event, the crew spotted the airfield late, and noticed that the spectators were lined up on the grass strip, not the paved strip on which they had been briefed. At this point the crew would have been well within their rights to abort the flypast altogether, and press on to the sightseeing trip over the Pyrenees which made up the latter part of their flightplan. Instead, the PIC elected to perform an unauthorised approach maneouvre to match the grass strip, even though they were considerably high and fast to do so.

HazelNuts39 16th Dec 2013 21:41


it maintained the airspeed determined by the approach settings.
I'm not sure what you mean by that. Are you referring to a point where the airspeed bottomed, changing from decelerating to accelerating while the engines continued to spool up from idle to TOGA?

DozyWannabe 16th Dec 2013 22:13


Originally Posted by HazelNuts39 (Post 8210490)
I'm not sure what you mean by that. Are you referring to a point where the airspeed bottomed, changing from decelerating to accelerating while the engines continued to spool up from idle to TOGA?

I'm referring to the constantly changing variable of airspeed throughout. The airspeed did not increase enough to have a significant impact on calculated maximum alpha until a point where they were just about to make contact with the trees. The engines spooled down well before the aircraft crossed the threshold. The aircraft maintained airspeed based on that initial value, and maintained pitch attitude equivalent to maximum AoA available. The airspeed didn't increase substantially until after the engines spooled up again.


All times are GMT. The time now is 04:57.


Copyright © 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.