Habsheim

[Chris Scott]

Confit,

I think you are wrong to argue that the wind was calm, if only for the reasons HN39 has pointed out.

Also, if you look again at the Andre Karsenty video, you will see a large balloon in the backround about 7 seconds before the a/c reaches the treeline. It is leaning slightly towards the south, suggesting some kind of headwind component for the A320.

And, as you and any pilot of any size of a/c know well, the proximity of trees is always associated with some degree of wind-shear.

Wind components (W/Cs) can, of course, be estimated from a comparison between the TAS and the GS. (Any error in the GS readings from the IRS are likely to remain constant for the short period of the straight-line flypast. In this case, allowing for pressure altitude and temperature, adding 3 kt to each IAS value gives the approximate TAS.)

Starting from level-off at TGEN 321 and finishing with TGEN 334, the estimated W/Cs at one-second intervals are as follows:
-7, -4, -8, -9, -7, -7, -6, -4, -6, -5, -7, -7, -3, -3.

The above variations are consistent with small gusts of a light headwind.

[CONF iture]

Originally Posted by HN39

That is not what I said in post #622 based on the graph in the report. In the simulation the airplane gained 10 ft and that is equivalent to 1 kt of airspeed.

Good ... then from the 30 ft with a gain of 10 ft at 17 deg of alpha and speed at 119 kt the BEA and yourself agree that the airplane was surviving the 40 ft forest ...

It is ridiculous to dismiss the simulation for no other reason than that you can't see a difference of one knot on that plot.

I am certainly not looking for any kind of "one knot difference" or other 2 kt gust theory to dismiss such ridiculous simulation when they want you to believe that taking 5 additional degrees of alpha in 2 seconds to bring it to alpha max won't make a difference in term of altitude and speed compared to the same aircraft that remains between 14 and 13 of alpha as computed on the graph.

Originally Posted by Chris Scott

I think you are wrong to argue that the wind was flat calm, if only for the reasons HN39 has pointed out.

Where did I argue so exactly ... ?
The atmosphere was calm, and the wind less than 5 kt according to the Air Traffic Controller.

Thanks for bringing to our attention that DFDR Tome 2 of the same document appears to show a figure of 34 ft, in line with the narrative. The version you show magnified in your post is, I presume, the PDF copy, the poor quality and assembly of which we discussed earlier on this thread.
Another curious aspect of this matter is that an HTML version of the BEA report - Habsheim F-GFKC - clearly shows a figure of 24 ft in both Tomes 1 and 2.

Here is the ultimate purpose of my concern. This Official Report has a typographic quality of the 60s at most but certainly not one of the 90s standards. All the "graph" as presented as they are, are a pure joke or an insult to the reader. Too convenient for blurring data and bringing confusion. One crucial Annexe is nowhere to be seen and manipulation in Annexe labeling is even created in an attempt to mask its absence. As the BEA is that unprofessional, we clearly do not need them.

I am concerned that a guy with your background seems to give more credibility to such piece of cr@p from the BEA than to Ray Davis ...

[gums]

dragging it out, heh?

C'mon, Confit.

The jet and its FCS worked as advertised. And then the laws of physics and aero came into play.

It could be that 'bus folks did not do enough testing at altitude to examine all the "protections" and "floors" and such using flight idle, TOGA, etc. The jet did not have the pitch coefficient problem we had in the Viper, so it had plenty of nose down authority at high AoA ( although we saw with AF447 that the THS position had to be compensated for to get the nose down). Nose up is "limited" by the flight control laws WRT AoA, not the actual capability of the plane, and we had that in the Viper. You can command everything, but HAL is only gonna let the jet "obey" the laws, best it can.

[noske]

Originally Posted by Chris Scott

Another curious aspect of this matter is that an HTML version of the BEA report - Habsheim F-GFKC - clearly shows a figure of 24 ft in both Tomes 1 and 2.

Well, that's the risk you take with an unofficial publication. That document was probably derived from the BEA PDF, and transcribing those DFDR printouts must have been especially error-prone. I'd say that the evidence offered by CONF iture (the magnified scans, and the values as they appear in the narrative) is sufficiently convincing.

[Chris Scott]

BEA Rapport Final

Quote from noske:
"Well, that's the risk you take with an unofficial publication. That document was probably derived from the BEA PDF, and transcribing those DFDR printouts must have been especially error-prone. I'd say that the evidence offered by CONF iture (the magnified scans, and the values as they appear in the narrative) is sufficiently convincing."

Yes, I agree with all of that. It is ironic, however, that the RA figure of 24 ft at TGEN 334 in the HTML version of Tome 2 is more plausible than the 34 ft figure in the PDF copy of the BEA report. The Karsenty video rules out any inference that the a/c might have climbed 10 ft in the last second.

Whichever of the two figures is correct, students of the report might expect that its final edition would have corrected an unnecessary anomaly that undermines the authority of the whole document. If that and other less imperative amendments were done, it is unfortunate that the BEA has not made the amended document accessible via their website.

In public transport, the Habsheim accident has unique aspects and complexities, and continues to provide a basis for the study of a broad range of aeronautics. In recent years, the BEA has published several very informative papers relating to flight safety matters in general. Its valuable contributions to the subject are somewhat undermined by the continuing deficiencies of the Habsheim document.

Hi gums,
Just a reminder to you and others that, in AF447, Pitch Alternate Law allowed the THS to continue auto-trimming to full nose-up trim. At Habsheim, starting with Pitch Normal law, the THS auto-trimming was disarmed first (briefly) by the premature, 2-second engagement of Flare mode at TGEN 313 (t -21) - triggered by flying over trees; and, secondly, when Flare mode was entered again at TGEN 317 (t -17). Finally, it remained disarmed when Alpha Protection mode took over from Flare mode at TGEN 330 (t -4). Earlier, in fact, it had had no need to trim between TGEN 306 (t -28) and t -21. So, to sum up, it remained at U04 (4 degrees nose-up trim) for about the last half a minute of flight. Full deflection in the nose-up trim sense is 13 degrees.

[roulishollandais]

In France the official version is not the BEA file, but the "JOURNAL OFFICIEL" copy of the BEA file. Probably the BEA had two versions : 24 and 34 Ft. They have to verify and ask to the J.O. to register the true number. The J.O. would not modify such a number.

[Chris Scott]

Bon soir roulis,

The Journal Officiel is the document, available via the BEA website, which Confit shows.

http://www.bea.aero/docspa/1988/f-kc...f-kc880626.pdf

Can you explain more? Are you saying that, once the document has been published as the Journal Officiel, the BEA is completely powerless to publish any necessary corrections?

[DozyWannabe]

The Ray Davis angle was covered earlier in the thread more than once - here:

http://www.pprune.org/tech-log/52803...ml#post8209613

and here (click the arrow to go to the original post):

Originally Posted by DozyWannabe

Notably, the ACI programme got this wrong too - it was not ATC's discrepancy with the CVR, it was a misinterpretation of the DFDR. [Davis] seems to have treated the DFDR's transmit/receive flag as referring to the call from ATC (which would be almost 5 seconds adrift from the ATC recording and unacceptable). In fact the DFDR does not flag incoming, only outgoing transmissions, and the DFDR referred to the crew's response (just shy of 1 second adrift from the ATC timestamp, and within the margin of error).

All this and more is in the Airbus document...

One thing that I didn't mention before and may have some bearing is that Mr. Davis was hired by a UK TV production company making a documentary on the subject - not of the crash itself, but of Asseline's attempt to clear his name. The Airbus document explaining where Davis may have erred does not allude to this, but there is a distinct possibility that the information supplied him may have been missing some of the details mentioned in the above posts.

We also know that the Loral DFDR units fitted to the early A320s were later known to be susceptible to write errors when subjected to unexpected shock or vibration. The estimated value of +0248 RA at TGEN335, which CONF iture has helpfully provided in magnified form, is clearly erroneous - therefore it is within the realms of possibility that the previous value is also in error. That the video evidence clearly does not demonstrate an altitude gain of 10ft before impact with the trees supports this hypothesis in my opinion.

Originally Posted by Chris Scott

Hi gums,
Just a reminder to you and others that, in AF447, Pitch Alternate Law allowed the THS to continue auto-trimming to full nose-up trim.

Hi Chris - I hope you find this interesting, though it's on a bit of a tangent. When we tested the AF447 scenario in an A320 sim (the best we could do as there were no A330 sims available), the autotrim was limited by the system well before we got into difficulties - to simulate the scenario, the TRE had to manually wind on full nose-up trim with the wheels. If such behaviour was consistent with the A320, it follows that autotrim behaviour on the A320 series is different from that on the A330/340.

[roulishollandais]

Originally Posted by Chris_Scott

Can you explain more? Are you saying that, once the document has been published as the Journal Officiel, the BEA is completely powerless to publish any necessary corrections?

Originally Posted by roulishollandais

In France the official version is not the BEA file, but the "JOURNAL OFFICIEL" copy of the BEA file. Probably the BEA had two versions : 24 and 34 Ft. They have to verify and ask to the J.O. to register the true number. The J.O. would not modify such a number.

@Chris Scott,
1. My last sentence was perhaps at a wrong place?
Once the document has been published in the Journal Official any necessary correction is possible in a similar form and level of law hierarchy, probably in a short but accurate description of the modification but with the signature of the original administration . If the modification is not published you may ask a Court to apply the wrong text. The Journal Officiel does the maximum to avoid such mistakes and many verifications are done and redone before publishing, to respect every comma or space, but in the JO typography, design , with absolute rules of registering, diffusion, existence of copy. It often needs much time which brings a delay between the BEA text who has its own design, and rules, published so quickly as possible and the JO tex
It is unusual that the BEA text provides a copy of the Journal Officiel which gets a BEA report.. Why, when was the original BEA report replaced in their files?

2. The numbers provided by the sensors must not be replaced, the BEA may only comment them in his report

[CONF iture]

Originally Posted by gums

The jet and its FCS worked as advertised.

Absolutely not - The advertisement is to do the necessary in order to expect alpha max 17.5 deg in this configuration, but certainly not to restrict it to 15 at most.

You can command everything, but HAL is only gonna let the jet "obey" the laws, best it can.

The law in this case is to permit 17.5 deg of alpha and not to restrict it to 15, or even 14 depending on where you look into that BEA Official Report ...

Originally Posted by noske

That document was probably derived from the BEA PDF, and transcribing those DFDR printouts must have been especially error-prone.

I'm not too sure how such document as you posted it is actually obtained ?
Don't you think the transcribing to obtain such format must be part of an automatic process ?

Originally Posted by Dozy

We also know that the Loral DFDR units fitted to the early A320s were later known to be susceptible to write errors when subjected to unexpected shock or vibration. The estimated value of +0248 RA at TGEN335, which CONF iture has helpfully provided in magnified form, is clearly erroneous - therefore it is within the realms of possibility that the previous value is also in error.

Then you should have said that that type of DFDR was also susceptible to write errors when anticipating a shock ... the emotion maybe.
BTW any data from TGEN335 has already no value has the tape is labelled as de synchronized.

That the video evidence clearly does not demonstrate an altitude gain of 10ft before impact with the trees supports this hypothesis in my opinion.

Not more than it demonstrates an altitude loss of 6 ft the second before ...

[DozyWannabe]

Originally Posted by CONF iture

The advertisement is to do the necessary in order to expect alpha max 17.5 deg in this configuration, but certainly not to restrict it to 15 at most.

Where does it say this? It doesn't say this in any of the documentation.

To my mind it's two separate pieces of information out of context that you have yourself combined to come up with this ludicrous theory that the FCS was to blame, and not the arrogant*, incompetent** and stubborn*** (on that day at least) Capt. Asseline.

* - In blithely assuming that he was skilled enough to take a greater level of risk with pax on board than Airbus's own test pilots
** - Thoroughly making a hash of the approach with very poor thrust management
*** - Going ahead with the flypast on the first attempt when it was obvious that the briefing was wrong (regarding which runway to follow)

[CONF iture]

Originally Posted by Dozy

Where does it say this? It doesn't say this in any of the documentation.

Of course it does as already posted in #599 and #603.

To my mind it's two separate pieces of information out of context that you have yourself combined to come up with this ludicrous theory that the FCS was to blame, and not the arrogant*, incompetent** and stubborn*** (on that day at least) Capt. Asseline.
* - In blithely assuming that he was skilled enough to take a greater level of risk with pax on board than Airbus's own test pilots
** - Thoroughly making a hash of the approach with very poor thrust management
*** - Going ahead with the flypast on the first attempt when it was obvious that the briefing was wrong (regarding which runway to follow)

Finally showing your true colors after initially playing false demagoguery by pretending "having some sympathy for Capt. Asseline".

[DozyWannabe]

Neither of those posts state 17.5deg as being a guaranteed swift response upon full application of back stick, and in fact the Airbus documentation does not mention 17.5 degrees at all. You're taking the absolute value given in the BEA report out of context and grafting it to the Airbus documentation in a way that is not consistent with the way the documentation is written in order to claim that there should have been an expectation of 17.5 degrees, when that doesn't seem to be the case.

And I *am* sympathetic to Asseline in some respects, particularly the lack of oversight and poor preparation materials given him by the airline. This does not alter the fact that the conduct of the flight seems to be consistent with the negative traits sometimes ascribed him, and *on that particular day*, as I said, he seemed to be displaying those traits during the conduct of the flight. The two positions are not mutually exclusive.

[HazelNuts39]

Originally Posted by DozyWannabe

That the video evidence clearly does not demonstrate an altitude gain of 10ft before impact with the trees supports this hypothesis in my opinion.

Originally Posted by CONF iture

Not more than it demonstrates an altitude loss of 6 ft the second before ...

The DFDR recording offers three sets of data that can be used to reconstruct the trajectory:

-- radio altitude
-- pressure altitude
-- x, y, and z accelerations

When comparing these three sources, the different heights of the respective sensors must be taken into account. The radio altimeter antenna is on the lower rear fuselage, the ambient pressure is sensed by Air Data Modules close to the static pressure ports on the front fuselage, and the three-axis accelerometer is located near the center of gravity. The relative height of the three sources varies with the pitch attitude of the airplane, and for the radio altitude also with the elevation of the terrain. The following graph shows the height of the CG taking those factors into account. The elevation of the terrain can be taken as 781 ft at TGEN=334 with a downhill slope of 0.15%. The pressure altitude depends on the local pressure, given in the report as QNH=1012 hPa and QFE=984 hPa. If the pressure has been accurately measured and properly rounded the QNH must then lie between 1012.1 hPa and 1012.5 hPa. The graph shows the CG height based on the pressure altitude for both values of QNH.

[gums]

Gotta tellya, Confit, that until I see the actual FCS charts/functions for modes and such, I'll stay with Doze and others that the plane gave everything it was "programmed " do.

My problem is with the "flare mode", which requires increasing back stick once the jet is "x" feet above the ground. Could it be that the "flare mode" prevents the jet from reaching the advertised max AoA?

I swear, the 'bus has more "laws" and functions and everything else that the space shuttle, the Viper or the F-22 do not have. For a pilot, I want to know exactly what to expect without having to go thru modes, sub-modes and then sub-sub-modes. Sheesh.

I am not an advocate of "direct" laws, as some have mentioned on our discussion. But you can only do so much to prevent/help some hamfist to fly the jet.

[DozyWannabe]

Originally Posted by gums

Could it be that the "flare mode" prevents the jet from reaching the advertised max AoA?

No - aircraft was not in flare mode from shortly (as in under a second) after it overflew the "copse"/"bosquet". Well out of flare mode when the trees were spotted, and as other A3x0 rated pilots have attested, flare mode can be cancelled simply with application of TOGA thrust*. As OG attested to (If I read him right, it's related to EFCS phugoid damping) - the Alpha Max command in High AoA Protection gives as much as it can, then likely waits a fraction of a second to stabilise before determining whether the remaining AoA is to be used for extra pitch or for bank input.

Oops - got the above regarding OG's post completely wrong - apologies!

EDIT : As I said earlier, I have a suspicion that the airspeed deltas have an effect on how long it takes to consider the attitude "stable", and in a case such as this one where the aircraft was decelerating almost right up to impact, the duration would likely be longer than if it had airspeed in reserve.

I swear, the 'bus has more "laws" and functions and everything else that the space shuttle, the Viper or the F-22 do not have.

The 'bus systems have three primary laws only. Alternate Law has a number of sub-modes, but the overall performance isn't a great deal different between them - the sub-mode is determined by which of the myriad systems has suffered a failure. It goes without saying that this will be more complex than a fighter, possibly even the F-22, because an airliner is a much more complex piece of engineering mechanically.

* - This aspect is discussed in more detail earlier in the thread.

[roulishollandais]

@Hazelnuts39
What is still unknown is the altitude read on the altimeter on the cockpit

[DozyWannabe]

Originally Posted by roulishollandais

@Hazelnuts39
What is still unknown is the altitude read on the altimeter on the cockpit

I think this has been covered before, but if I may attempt to recap...

The only way to be 100% certain would be to have had a camera trained on the PFD, but since we don't have that, a degree of assumption will always be necessary. Folks, correct me if I'm wrong - but I think from recollection of the AF447 thread that the DFDR altimeter trace is fed from the ADIRU selected on the Captain's side (LHS).

As others have demonstrated, the DFDR trace seems to tally well with other data when it comes to actual baro altitude. QNH seems to have been correctly set, and all things considered the likelihood is that the baro alt on the PFD was showing the values recorded on the DFDR. For there to have been a discrepancy would mean a fault on the data bus between the ADIRU and the LHS PFD, for which no evidence seems to exist (though, admittedly, such evidence would be difficult to trace later on).

Asseline insists that the baro alt was reading 100ft throughout, and furthermore frames the possibility that it wasn't as tantamount to accusing him of lying. He mentions earlier baro alt misreadings on test flights and argues that may have happened in this case. I believe he states elsewhere that he was relying on the baro alt and external references because he found the digital RA display difficult to read.

I'm sure that Asseline believes fervently that the baro alt was reading 100ft, and if it wasn't, he may have still perceived it as such through confirmation bias - as the workload on the flight deck was significantly higher than anyone had anticipated due to the briefing errors. If the baro alt was in fact consistent with the DFDR, that doesn't make him dishonest - it just raises the possibility that he made a (completely understandable) mistake under pressure.

On the other hand, his own input raises further questions about the conduct of the flight as a whole - particularly with regard to the wisdom of relying solely on one instrument if you consider that instrument to be unreliable, with the only safeguard being external references at an unfamiliar airfield.

[CONF iture]

Originally Posted by Dozy

You're taking the absolute value given in the BEA report out of context and grafting it to the Airbus documentation in a way that is not consistent with the way the documentation is written in order to claim that there should have been an expectation of 17.5 degrees, when that doesn't seem to be the case.

As 17.5 deg value for alpha max at CONF3 given by the BEA is not good enough for you ... where do you think the BEA got that value if not from Airbus ?
True I should always remember that you still don't know which config was used in Habsheim ... !?

As OG attested to - the Alpha Max command in High AoA Protection gives as much as it can, then likely waits a fraction of a second to stabilise before determining whether the remaining AoA is to be used for extra pitch or for bank input.

Another nice piece of disinformation. Who should I credit it to, OG or you ... ?

As I said earlier, I have a suspicion that the airspeed deltas have an effect on how long it takes to consider the attitude "stable", and in a case such as this one where the aircraft was decelerating almost right up to impact, the duration would likely be longer than if it had airspeed in reserve.

Then Bechet in his simulator just proved your suspicion wrong ...

Originally Posted by gums

Gotta tellya, Confit, that until I see the actual FCS charts/functions for modes and such, I'll stay with Doze and others that the plane gave everything it was "programmed " do.

Sorry gums but I go by the book, now if the the plane behaved as per an unpublished program ... it was the BEA duty to detail it afterwords.

[DozyWannabe]

Originally Posted by CONF iture

As 17.5 deg value for alpha max at CONF3 given by the BEA is not good enough for you ... where do you think the BEA got that value if not from Airbus ?
True I should always remember that you still don't know which config was used in Habsheim ... !?

If you say it was CONF 3 and the report says it was CONF 3, then it was CONF 3. I've been aware of that for a while.

Of course the BEA would have got the value from Airbus. But the issue I have with your position regarding 17.5deg is not the value itself - it's your interpretation of combining the absolute value as given in the BEA report with the less-specific (regarding values and how they are used by the EFCS) documentation that Airbus provided to flight crew.

Your argument seems to be that Airbus - to use your own word - "advertised" that in CONF 3, 17.5 degrees AoA would definitely be achieved under any circumstance with the application of full back-stick, when the documentation says no such thing.

AZR said much the same thing to you a couple of pages back:

Originally Posted by AlphaZuluRomeo

From an aerodynamic point of view, there were probably more than 2.5 deg before stall. From an FCS point of view, for this to be correct, the FCOM should read that alphamax was to be attained immediately and without any damping.
Does the BEA report or the FCOM read this? No. You just quoted both.
The NTSB report on Hudson event, and numerous other discussions here and there have put forward several explanations as to why alphamax(17.5) would, in certain circonstances, not be reached immediately.

Another nice piece of disinformation.

On what are you basing that accusation? There seems to be a definite short delay between acquiring 15 degrees AoA and going further. OG reckoned it was related to damping.

Then Bechet in his simulator just proved your suspicion wrong ...

How so? Please elaborate. For one thing, you yourself said that Bechet's work in the simulator and real-world testing did not involve a precise reconstruction of the event - so I'm puzzled as to how they could disprove an assertion if that was the case.