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

Habsheim

Old 30th Mar 2014, 20:28
  #721 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Originally Posted by CONF iture View Post
...and a stall could not have happened after clearing the trees as alpha max is not alpha stall.
You clearly didn't read what I wrote, because I said they were comparing it to a *conventionally-controlled* aircraft. Alpha Max is irrelevant in that situation because full back-stick/yoke does not command Alpha Max in a conventional aircraft, it just moves the elevators to the stop - if the aircraft is moving fast enough, it will pitch up and climb, but if it is not, it will pitch up and stall.

Good for you ... what I was expecting is a BEA + Airbus analysis.
By the way if phugoid damping was involved, they could have detailed why Bechet was miraculously spared from it ... ?
As HN39 said, phugoid damping is not used when in pitch command (i.e. the mode Bechet was in), only in HAP mode (which Asseline was in).

As I said before, the fact that the investigation went as far as confirming the behaviour as normal, but no further, could be argued as a point that they missed an opportunity to further explore the mode switches (and consequences of those changes) involved. However, the events of the last three seconds in the context of an accident investigation could equally be considered a minor point when compared to the major systemic failures that led to the aircraft being put in that position. Namely:
  • The airline provided woefully inadequate briefing materials
  • The airline failed to perform even a cursory review and risk assessment of the flightplan
  • The airline failed to check the flightplan against DGAC regs regarding minimum altitude for display flights
  • The airline did not require the crew to submit an effective "Plan B" for each stage of the flight plan
  • The airline did not require the flight crew to familiarise themselves with the airfield, other than how it appeared on the chart (i.e. no prior recce flight was required)
  • Neither the airline nor the flight crew attempted to confirm which runway was in use for the airshow on that day

The result of these failures was that the flight crew were effectively entirely reliant on their ability to improvise at short notice - which, given the inherent riskiness of the manoeuvre, should have been completely unacceptable on the face of it. Even more so when you take into account that there were pax on board.

To the best of my knowledge, "Priority One" of accident investigation is working to ensure that the incident is never repeated.

Now, I'm no accident investigator, but if I were I'd be inclined to consider getting to the bottom of how the aircraft was permitted to get into that position in the first place as a much higher priority than getting too far into why the Alpha Max command was complied with more slowly than it otherwise might have been (especially given the fact that the flight had been woefully mismanaged long before Asseline pulled that stick back).

Put another way, if the mistakes and loopholes in the list above can be prevented and closed, then with all other things considered there should never be another aircraft put in that position again, making the assessment of precisely why the aircraft was slow to achieve Alpha Max somewhat of a moot point.

Again, I said before that Asseline's best hope of being able to get the investigation to look further into the matter of what the EFCS was doing in those last few seconds would have been to take his lumps and remain positively engaged with the investigation throughout. By disengaging from the investigation and subsequently actively briefing against it, he did his own cause more harm than good.

Last edited by DozyWannabe; 30th Mar 2014 at 20:39.
DozyWannabe is offline  
Old 30th Mar 2014, 21:27
  #722 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Originally Posted by OK465 View Post
If I'm reading it correctly, I am guessing that what you're referring to here is the fact that in the simulation, the elevator smoothly continues to trend leading edge less down (less negative)/aircraft less nose up until the end of the plot.
Now look at the thrust and IAS, and what the pitch is actually doing.

Last edited by DozyWannabe; 30th Mar 2014 at 22:08.
DozyWannabe is offline  
Old 30th Mar 2014, 22:47
  #723 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 756
metalanguage

Metalanguage = object formal language in semantic brackets.

Example of semantic brackets in character strings is using a special character which is never used in the object language.
As the number of characters is very limited, the special character is often an unused string of characters, like : = , http://www , Ctrl Z, [/quote] and [quote], etc.

Building an object language vs metalanguage may use different gramatic too but it is not mandatory.

Children playing to be master or parents or king mostly build metalanguage to be considered as master, parents or king! Sometimes they forget to use the metalanguage and suddenly they are no more king, and everybody is laughing.

In the simulator a hidden mistake (forget the space between ":" and "=" and you discover a smiley no-no instead of the string ": =" ) in the metalanguage semantic brackets may modify the behaviour of the simulated system (plane, inertial system, aso) without you know it, and you have a bad surprise when you do the same action on the plane, despite your aerodynamic algorithm is OK.

It is always difficult to test the semantic brackets, you would mostly need a meta-meta-language!

All that applies well on well structured simple systems using formal language. Graceful degradation brings a greater complexity .... KISS!

Last edited by roulishollandais; 30th Mar 2014 at 23:16. Reason: spaces lost in copying
roulishollandais is offline  
Old 30th Mar 2014, 23:05
  #724 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
@roulis - I've said before, the computer code is not manually keyed in in that manner. Code blocks are tested in situ against the formalised function, then assigned to an element in a graphical development environment, which then generates the logic path from those individually-tested blocks of code. Each subsystem is then itself tested against the required behaviour pattern, and as the subsystems are linked together to form an overall program "module", that module and its functions are tested individually and in conjunction with others.

These tests are run *every single time* a change is made, and the results of the tests plotted against a regression graph. Any change in that graph is immediately apparent, and will point to exactly where the program deviates from the specified behaviour.
DozyWannabe is offline  
Old 30th Mar 2014, 23:33
  #725 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
You clearly didn't read what I wrote, because I said they were comparing it to a *conventionally-controlled* aircraft.
Typical Dozy disinformation again - where is the quote ?

As HN39 said, phugoid damping is not used when in pitch command (i.e. the mode Bechet was in), only in HAP mode (which Asseline was in).
Speculation only from HN39 that would demonstrate how the system is inefficient but you're happy to take it for cash ... Where is the official reference ?
CONF iture is offline  
Old 30th Mar 2014, 23:52
  #726 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Originally Posted by CONF iture View Post
Typical Dozy disinformation again - where is the quote ?
Here's a novel idea, why don't you find it and prove me wrong?

EDIT : Certainly whenever I've heard that factoid referred to in the past, I got the impression that they were talking about an unprotected (i.e. conventional) aircraft - not least because I can't see how it makes sense any other way!

Speculation only from HN39 that would demonstrate how the system is inefficient but you're happy to take it for cash ... Where is the official reference ?
It's not inefficient, it's a necessary consequence of an aircraft being commanded to achieve or hold a specific AoA - namely a tendency to begin a phugoid motion.

On a different subject, I strongly recommend you have a look at OG's post #734, then look at the graphic on Annex 10 p.13. Bear in mind that what you're looking at (I think) is a simulation of how the aircraft and systems would have responded to the commands as they were prior to impact had there been no collision (114 kt being very close to the airspeed achieved just prior to impact).

Last edited by DozyWannabe; 31st Mar 2014 at 00:49.
DozyWannabe is offline  
Old 31st Mar 2014, 00:50
  #727 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by Dozy
Here's a novel idea, why don't you find it and prove me wrong?
Assume the disinformation you're spreading around or retract it.

It's not inefficient, it's a necessary consequence of an aircraft being commanded to achieve or hold a specific AoA - namely a tendency to begin a phugoid motion.
Knowing both are starting from identical parameters, why Asseline should be subject to it but Bechet not ... ?
CONF iture is offline  
Old 31st Mar 2014, 01:02
  #728 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Originally Posted by CONF iture View Post
Assume the disinformation you're spreading around or retract it.
Why?

(Or more to the point, how is it fair that you get to make unsubstantiated claims of a "restriction" in the EFCS (and insinuate that it may have been surreptitiously removed) unchallenged, whereas I must retract any statement with which you disagree out of hand?)

Knowing both are starting from identical parameters, why Asseline should be subject to it but Bechet not ... ?
They're not starting from identical parameters. As stated:

The difference is in pulling the sidestick back - Bechet at the same instant and Asseline 3 seconds later. So Bechet was still in pitch control law when he pulled the stick back while Asseline was in alpha-prot.
Check the graphs on Annexe X, page 9. Look at the bottom of the page (when rotated lengthwise), and you'll see a graph indicated "Manche Longitudinal". Asseline's inputs are the broken, dotted plot and Bechet's are the unbroken plot. Note that on this particular run, Bechet snaps the stick back almost immediately approximately 3 seconds prior to when Asseline did the same. Bechet's action is just prior to when the mode switched from pitch command to HAP mode. Asseline's motion is also more gradual.

Last edited by DozyWannabe; 31st Mar 2014 at 01:33.
DozyWannabe is offline  
Old 31st Mar 2014, 02:02
  #729 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by Dozy
Why?
That's what responsible adults would do.

They're not starting from identical parameters.
Yes they are, according to the graph at the remise des gaz, except that Bechet is 10ft lower ...

Note that on this particular run, Bechet snaps the stick back almost immediately approximately 3 seconds prior to when Asseline did the same. Bechet's action is just prior to when the mode switched from pitch command to HAP mode. Asseline's motion is also more gradual.
Then if I had to guess the one who could be more prone to phugoid motion I would go for Bechet not Asseline.
CONF iture is offline  
Old 31st Mar 2014, 02:09
  #730 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Originally Posted by CONF iture View Post
according to the graph at the remise des gaz, except that Bechet is 10ft lower ...
Not quite 10ft, and Bechet pulls back roughly at the same point as "remise des gaz", whereas Asseline waits for around 3s before doing the same.

Then if I had to guess the one who could be more prone to phugoid motion I would go for Bechet not Asseline.
How? Asseline was in HAP mode when he pulled full-aft, Bechet was in pitch command.
DozyWannabe is offline  
Old 31st Mar 2014, 02:55
  #731 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by Dozy
Not quite 10ft
How much then ?

How? Asseline was in HAP mode when he pulled full-aft, Bechet was in pitch command.
It does not change anything - they still have to deal with the same level of energy.
Still waiting for the reference stating phugoid damping under high AoA protection but none under pitch command ...
CONF iture is offline  
Old 31st Mar 2014, 08:52
  #732 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 80
Posts: 1,689
Originally Posted by CONF iture
Speculation only from HN39
Originally Posted by CONF iture
Still waiting for the reference stating phugoid damping under high AoA protection but none under pitch command ...
Originally Posted by HN39 #707
Phugoid damping is a feature of the alpha-protection mode, because an airplane is prone to develop phugoid motions when constrained to a commanded angle of attack. Flying an airplane to a commanded pitch attitude does not result in phugoid motion, and therefore pitch-command mode does not require phugoid damping.
Firstly, both the Bilbao report and the Hudson river report mention phugoid dammping explicitly for the high angle of attack protection mode. If you have a reference stating that it is a feature of the pitch command mode then please provide that reference.

Secondly, I explained in post #707 why phugoid damping is necessary in alpha-prot and unnecessary in pitch command mode. I showed a phugoid motion that results from maintaining AoA after a disturbance of a steady flight condition. On the AF447 thread I showed similar simulations of airplane motion at constant pitch attitude after a pitch change, for example:

Last edited by HazelNuts39; 31st Mar 2014 at 13:41. Reason: Graph added
HazelNuts39 is offline  
Old 31st Mar 2014, 15:04
  #733 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by HN39
If you have a reference stating that it is a feature of the pitch command mode then please provide that reference.
Not my point - What makes phugoid damping a feature of high AoA protection but irrelevant when high AoA protection is activated from pitch command initially ?
In other words, by which subterfuge Bechet is not subject to phugoid damping as he is obviously hitting high AoA protection too ?

The point has been brought here but you did not comment.
CONF iture is offline  
Old 31st Mar 2014, 15:30
  #734 (permalink)  
 
Join Date: Jun 2011
Location: West of Offa's dyke
Age: 84
Posts: 474
OK465

What's at play here? Is this solely due to the engine's increasing thrust pitching moment?
That's the way I read it. A quick sum suggests that at 118kt (the end of the simulator run) 2 CFM56s at full chat would give a nose up pitch equivalent to about 10 deg up elevator.

Why wouldn't the elevator show reversals of direction or at least pauses during the pitching to stabilize AOA, if AOA is truly the driver?
For the reason given above, the net pitching moment (thrust plus elevator) at the end of the run was equivalent to about 7 deg nose up. There is an indication that the elevator is starting to move back up at the end of the run, but even at that point the pitch is barely sufficient to be consistent with the commanded alphamax of 17.5 deg plus a bit of climb. [The pitch scale appears to be incorrectly labelled which doesn't help!] It looks to me as if for most of the time the thrust is supplying more than enough pitch to satisfy the demand but towards the end the thrust contribution is levelling off and it is from that point onwards that one would see elevator movement alone doing the controlling/stabilising job. Unfortunately (for this discussion at least) there isn't enough time in that state to form a judgement.
Owain Glyndwr is offline  
Old 31st Mar 2014, 15:40
  #735 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 80
Posts: 1,689
Originally Posted by OK465
BTW I notice on HN39's presented graph that what is referred to as alphamax is actually higher than alpha SW, I assume that reference is to alphaCLmax and not the conceptual protection alphamax which is lower than alpha SW.
For the A330 in clean configuration alphamax (normal law) is greater than alpha_SW (alternate law) at all Mach numbers less than 0.86. I do not know the value of alphaCLmax, except that it is greater than alphamax as explained in numerous Airbus documents. At high Mach number alphamax is actually buffet onset, as explained in the FCOM.

P.S.
I forgot to mention that we know one point of the alphaCLmax vs Mach curve: AF447 stalled at 10 AoA / M.67.

Last edited by HazelNuts39; 31st Mar 2014 at 19:26. Reason: P.S.
HazelNuts39 is offline  
Old 31st Mar 2014, 16:10
  #736 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 80
Posts: 1,689
Originally Posted by CONF iture
Not my point - What makes phugoid damping a feature of high AoA protection but irrelevant when high AoA protection is activated from pitch command initially ?
In other words, by which subterfuge Bechet is not subject to phugoid damping as he is obviously hitting high AoA protection too ?
"Not my point"? IIRC you made it a point.

Bechet pulls the sidestick abruptly fully back at an AoA where the FCS is in pitch command mode and there is no phugoid damping. Therefore the airplane starts to pitch up achieving a rate of 2.5/second one second later. About again one second later the AoA exceeds 14.5 and the FCS goes into alpha-protection mode. The elevator then moves to more nose-down positions in a way that is quite similar to the accident sequence and the rate of rotation starts to decrease. However, the inertia of the airplane pitching up at 2.5/second and the nose-up pitching moment of the thrust increase allow the airplane to continue pitching up to about 16.5 while it begins to climb. That is what I see in that plot, although the resolution of the plot is not sufficient to show the AoA that is achieved.

Last edited by HazelNuts39; 31st Mar 2014 at 17:42.
HazelNuts39 is offline  
Old 31st Mar 2014, 17:49
  #737 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by HN39
"Not my point"? IIRC you made it a point.
My point was more precise and needed to be detailed :
What makes phugoid damping a feature of high AoA protection but irrelevant when high AoA protection is activated from pitch command initially ?
In other words, by which subterfuge Bechet is not subject to phugoid damping as he is obviously hitting high AoA protection too ?
Originally Posted by HN39
However, the inertia of the airplane pitching up at 2.5/second and the nose-up pitching moment of the thrust increase allow the airplane to continue pitching up a few degrees while it begins to climb.
No stall - Alpha max - How great is it really !
You just confirm here that everything was in the aerodynamics but only the FCS restricted Asseline at 2.5 deg short of alpha max and deprived him the opportunity to survive the poor situation he initially put himself in.
CONF iture is offline  
Old 31st Mar 2014, 18:20
  #738 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Originally Posted by CONF iture View Post
but only the FCS restricted Asseline...
Why do you keep calling it a "restriction"? It isn't - it's a deliberately engineered damping feature to ensure stability when in HAP mode which is, after all, concerned with protection from stall and stabilising trajectory.

Asseline wasn't restricted by the EFCS, he was restricted by his own actions in the previous 30 seconds or so - chopping back too much thrust, and then seeming indecisive about whether to actually go for Alpha Max (while constantly bleeding off airspeed). The EFCS doesn't know there's an obstacle three seconds away, it just tries to comply with the commands as best it can, and in this case both the TOGA thrust and full back-stick came too late.

The A320 and its systems were designed for ferrying passengers between airports, and the protections were designed to fit that function - it was not designed for "playing chicken" with forests at low altitude and low speed!
DozyWannabe is offline  
Old 31st Mar 2014, 19:07
  #739 (permalink)  
 
Join Date: Jun 2011
Location: West of Offa's dyke
Age: 84
Posts: 474
OK465

If I'm reading it correctly, I am guessing that what you're referring to here is the fact that in the simulation, the elevator smoothly continues to trend leading edge less down (less negative)/aircraft less nose up until the end of the plot.
Excuse my jumping back a couple of posts, but what I was actually describing as interesting was, starting from 114 kt, the sequential increases in thrust, airspeed, pitch and height. Each lags its predecessor as it must, but the end point is that the simulator run showed that the aircraft would have got to alphamax by about TGEN 336 and (with a reasonable extrapolation) 40 ft about two seconds later. This would be an elapsed time of around 7 or 8 seconds from the application of maximum back sidestick. The two other simulator exercises, Bechet's and an un-named pilot's show very similar times.

If we accept (as I think you do) that this simulator was a 'high fidelity' representation of the aircraft then these figures must mean something in terms of the actual aircraft behaviour on the day of accident. One might conclude that the EFCS was carrying out the pilot's command, but that the physics of the flight conditions were such that it would have needed around 7 or 8 seconds to complete the task.
Owain Glyndwr is offline  
Old 31st Mar 2014, 19:49
  #740 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 80
Posts: 1,689
Originally Posted by CONF iture
You just confirm here...
That is a strange twist - many posts ago I first suggested that phugoid damping was a possible explanation for the 'attenuated' response of the airplane to full back sidestick.

Last edited by HazelNuts39; 31st Mar 2014 at 21:15.
HazelNuts39 is offline  

Thread Tools
Search this Thread

Contact Us Archive Advertising Cookie Policy Privacy Statement Terms of Service

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