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

AF 447 Thread No. 5

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

AF 447 Thread No. 5

Old 8th Aug 2011, 14:40
  #1761 (permalink)  
 
Join Date: Nov 2007
Location: Scotland
Posts: 21
Likes: 0
Received 0 Likes on 0 Posts
stall indicator

I'm having a problem keeping up with the volume of posts, but I don't expect anyone has suggested this:
Hang glider pilots often tie a ribbon to the down-tube.
Maybe something that could be seen from the cockpit, but slightly more robust than a ribbon could work?
Bill G Kerr is offline  
Old 8th Aug 2011, 14:54
  #1762 (permalink)  
 
Join Date: Jan 2008
Location: uk
Posts: 857
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by Smilin_Ed
Dozy, it you have relevant arguments on why autotrim should not drop out with the autopilot, I will be glad to consider them.
Can't speak for Dozy, but two relevant arguments that come to mind are Schipol and Bournemouth. The 737 autotrim does drop out with a/p - and the consequences when pilots didn't immediately realise weren't pretty (although at Schipol it was probably a minor contributor with not watching the thrust the major one).

The common thread to me is not that A control laws are better than B, or visa versa, but that pilots should be properly trained to fly the a/c they are flying, including stall recognition and recovery. It appears that this has been lacking across the industry for a while, hopefully now in the process of being fixed.
infrequentflyer789 is offline  
Old 8th Aug 2011, 15:06
  #1763 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by Bill G Kerr
I'm having a problem keeping up with the volume of posts, but I don't expect anyone has suggested this:
Hang glider pilots often tie a ribbon to the down-tube.
Maybe something that could be seen from the cockpit, but slightly more robust than a ribbon could work?
Bill,
Something "slightly more robust" already exists. It's known as an AoA vane (sensor). There are a couple of those fitted to the nose.
Your ribbon is no more a 'stall indicator' than the AoA vane (or indicator), unless you know the stall angle of attack....
And there is no indication so far that the AoA sensors had any problems (unlike Perpignan).
ChristiaanJ is offline  
Old 8th Aug 2011, 15:22
  #1764 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,916
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by ChristiaanJ
And there is no indication so far that the AoA sensors had any problems (unlike Perpignan).
But what about number 1 stuck at 2.1 deg ?
CONF iture is offline  
Old 8th Aug 2011, 15:23
  #1765 (permalink)  
 
Join Date: Jan 2008
Location: uk
Posts: 857
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by alainthailande
Been a systems engineer for almost 30 years now, and I've never heard of a signed 16-bit implementation different from the traditional -32768 to 32767 that has exactly one representation for zero (what you call "minus zero" actually being -32768). Any reference to this particular implementation you're mentioning?

Old engineer as well.
Ones complement and sign-magnitude both do this.

Not sure how much either is still in use though - not a lot, most things are twos complement, but maybe in some niches.

I have a very vague (and probably wrong) recollection that fixed-point datatypes in Coral66 and maybe ADA (not used that much) might have used ones-complement magnitude under the hood. On some hardware. Coral was definitely used in aviation (maybe still is, I haven't worked with it since the 80s) as is ADA.
infrequentflyer789 is offline  
Old 8th Aug 2011, 15:31
  #1766 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by CONF iture
But what about number 1 stuck at 2.1 deg ?
Missed that, you got a link?
ChristiaanJ is offline  
Old 8th Aug 2011, 15:51
  #1767 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Likes: 0
Received 0 Likes on 0 Posts
Hi Christiaan,
Originally Posted by ChristiaanJ
Missed that, you got a link?
Yes, Confiture is right.
See for example longitudinal axis simulation graphs p.44, incidence notes on detailed flight parameters p.30-33 (French report), as well as other graphs.

Below, "incidence IRS1 [DA] (FDR)" is showing a value for alpha probe #1 which remains blocked at 2.1 until after 0210:50; later, it follows other probes but its output is under-reading.

takata is offline  
Old 8th Aug 2011, 16:43
  #1768 (permalink)  
 
Join Date: May 2000
Location: Glorious West Sussex
Age: 76
Posts: 1,020
Likes: 0
Received 0 Likes on 0 Posts
Smilin-Ed
Autotrim has its place, but not in my cockpit when I am trying to hand-fly.
Only one question: In a FBW Airbus how will you as pilot sense which way to trim?
TyroPicard is offline  
Old 8th Aug 2011, 16:55
  #1769 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes on 0 Posts
Explain why Trim needed when trying to recover, Loss of Control? Autotrim is like the co-pilot moving the wheel, but in secret? No bicycle bell, or Clacking?
Lyman is offline  
Old 8th Aug 2011, 17:04
  #1770 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes on 0 Posts
takata,
Merci pour le lien.
I admit I missed that.
Not sure from where the FDR takes the "incidence IRS1 [DA] (FDR)" signal (I somehow doubt it's the 'raw' AoA vane signal), but I agree it further complicates the issue.....
ChristiaanJ is offline  
Old 8th Aug 2011, 17:15
  #1771 (permalink)  
 
Join Date: Aug 2007
Location: London, New York, Paris, Moscow.
Posts: 3,632
Likes: 0
Received 0 Likes on 0 Posts
but I agree it further complicates the issue.
Perhaps.
Perhaps it points towards the amount of ice on the airframe??
Perhaps not.
glad rag is offline  
Old 8th Aug 2011, 17:53
  #1772 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,617
Received 68 Likes on 23 Posts
Autotrim, THS. "hand flying"

Many previous planes had "beeps" or other indications that the autopilot or the "autotrim" ( hate that word) was moving the HS. In fact, an infamous crash near Indianapolis caused by LOC and ice building on the HS has a "whoof" sound on the CVR indicating that "otto" was trimming. When the crew disconnected the AP, shazam! Other big time problems with that incident, for another thread.

The huge HS can easily override the relatively small movements of the actual elevators. Just think about the area and remember the real "law" about rho vee squared times Cl times surface area. Then there's ol' mach. So near the critical mach, the elevators at the rear of the HS can become ineffective, hence the "all moving" stab on all fighters and some commercial planes since 1950.

The problem we dinosaurs have with some FBW implementations ain't the fact that electrical signals go to the actuators versus cables, pushrods, or hydraulic lines. Pure hydraulic pressure controlled from the stick/yoke valves that have been around for 60 years. The problem is that the engineers and maybe some pilots added "features", "protections", etc. that get in the way of hand flying when various components so to lala land. No commercial plane I know of is certified if it's basic aerodynamics are not up to spec.

Because we dinosaurs are used to trimming the pressures off of the stick/wheel/yoke, we expect things to remain relatively calm after a bump in the air or completely letting go of the controls. And if we were using otto, we expected some indication that otto had trimmed the plane versus us. So any reversion from automated flight to manual flight should not require instantaneous analysis of what the hell is going on.

I repeat for the nth time........ You can build and implement a FBW airplane that flies exactly like those of yore. Replace the tubes, cables, etc. with electrical wires and no big deal. It's only when many "features" and "protections" are added that you have problems with the training and competence of the crew. I only flew one state-of-the-art high performance jet that had a decent artificial feel to the stick, and it was a bellows doofer connected to the stick thru a pneumatic line that made it harder to pull/push when the airspeed went up. So the idea of force feedback to the FBW stick would be nice, but not completely necessary. In fact, it would be veeeeery nice as long as we had an indication other than our seat-of-pants that it was inop.

The basic rules are still applicable when a lotta electronics and such go away - pitch and power for last known condition, then go as gently as possible from there.
gums is offline  
Old 8th Aug 2011, 18:52
  #1773 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,916
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by glad rag
Perhaps it points towards the amount of ice on the airframe??
But Ice Detectors 1+2 didn't detect any (P110 EN)

Originally Posted by gums
You can build and implement a FBW airplane that flies exactly like those of yore. Replace the tubes, cables, etc. with electrical wires and no big deal. It's only when many "features" and "protections" are added that you have problems with the training and competence of the crew.
... but also problems with sensors and probes and everything that has to process their readings ...

In the meantime, when everything is by the book, it is Fantastic !


Originally Posted by ChristiaanJ
I admit I missed that.
Question is why BEA 'missed' it too ... ?
CONF iture is offline  
Old 8th Aug 2011, 18:56
  #1774 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by gums
And if we were using otto, we expected some indication that otto had trimmed the plane versus us. So any reversion from automated flight to manual flight should not require instantaneous analysis of what the hell is going on.
Look at the graphs cyan curves I posted above: graph 6 "Position PHR" = THS setting, and graph 7 "Position gouverne de profondeur gauche" = Left Elevator setting. Now, compare them.

As expected, THS is not following short term g-load elevator demands. During the whole sequence, 0209:40 - 0210:54, THS setting is changing from -2.8 to -3.4 (negative = NU)... hence a maximum variation of less than 0.6 in 1.14 minute. During this whole sequence, the long term pitch demand increased from about 3 NU to 11NU, with up and down in between.

Stall warnings, sounded at 0210:51 and stall followed 10 seconds later. From this point, the large NU demand was confirmed both in amplitude and in duration by elevator ncrease up to Max NU deflection, and they were mostly maintained here during the following minute. Consequently, THS followed its order by moving in the very same way up to Max NU deflection.

There is absolutely no mystery about that: OTTO had NOT trimmed the plane versus pilot demand !!
- OTTO only did what it was asked to do!
Of course, this end result, with hindsight, is contrary to the good setting for a rapid recovery action... but nobody really wanted to recover from this situation at the first place!
takata is offline  
Old 8th Aug 2011, 19:30
  #1775 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes on 0 Posts
If this pilot is pulling, He has help from this THS. So he must take the timing of his pull, rmember how long, and hold down the same amount when he wants the nose to go lower? This pilot has not remembered some things, how will he know what to push? Is pushing also harder than pulling now, it gets less nose move?

You say loudly the trim is because of the pilot, but not the other? The pilot is not doing what he is to do? He does not know this trim, even if he remembers the airplane actions from gthe training? The THS knows, why not tell the pilot also?

What, you say noone wants the recovery? Also that the airplane in this shape is too slow for r4ecovery? What is this?
Lyman is offline  
Old 8th Aug 2011, 21:09
  #1776 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by infrequentflyer789
Did the artificial feel include airspeed as input ? - intuitively I feel it must do (and I believe B777 does), but deltas are very different so maybe it wasn't needed ?
Sorry, I'll have to pull out the doc... artificial feel was not "my" system. But yes, I agree, I would think "q" was definitely one of the inputs.

If so, what did the system use in the event of uAS ?
I can't remember any UAS incidents ever being mentioned.

On a vaguely related note, did Concorde ever suffer from pitot icing / similar uAS or at FL60+ was it simply not a problem ?
Concorde went above Mach 1 at about FL350. At cruise altitude (FL500 to FL600) and Mach 2, TAT was in the order of +120C, so icing was not really a problem then. And yes, at that altitude, Concorde was basically 'above' the weather, including the ITCZ..
ChristiaanJ is offline  
Old 8th Aug 2011, 21:58
  #1777 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,617
Received 68 Likes on 23 Posts
AP disconnect and autotrim

You are correct, TK.

And seems to me that the 'bus AP disconnect is very benign.

The problem we dinosaurs have is we still see things like "overspeed/mach" inputs and such versus simply letting the crew fly a fairly good aero design with appropriate warnings that they are going too fast, or getting too slow, or exceeding some AoA values, and so forth.

CONF has brought up a good point. My discourse on the FBW design concepts was to illustrate that we don't need an awful lot of sensors to fly a plane with a good, basic aero design. After all, we're not talking the F-22 or Eurofighter or Viper. Basic aero laws will provide static and lateral stability. But if we fool with "Mother Nature" by flying with an aft c.g., we can get into trouble. For the AF447 incident, looks like c.g. was not as far aft as would be normal for that phase of flight, so I rule out a serious static stability problem due to the c.g. One of the only reasons we didn't have a "direct" law in the Viper was the thing was intentionally designed and flown with negative static stability until above 0.9 mach or so. During the flyoff with the YF-17, the performance difference was dramatic. You can still see the difference when the Blues perform, or a Super Bug does a demo flight. That critter is very comfortable at extreme AoA because it is a conventional design. End of war story.
gums is offline  
Old 8th Aug 2011, 22:05
  #1778 (permalink)  
 
Join Date: Aug 2011
Location: Near LHR
Age: 58
Posts: 37
Likes: 0
Received 0 Likes on 0 Posts
That AoA #1 reading...

{New here - Hi to all }

That initially "stuck" reading (according to the FDR) from AoA sensor #1 is interesting, isn't it. Presumably its output got voted out of being used, due to the discrepancy with the other two sensors, but the fact that it was different doesn't seem to be explained in the BEA report.

Originally Posted by CONF iture
But Ice Detectors 1+2 didn't detect any (P110 EN)
Understood, but that lack of report from the detectors, doesn't prove ice wasn't present ("absence of evidence" and all that...), especially since there seems to be evidence of ice in that CVR noise noted by the BEA (page 73 English version).

What else could be the cause? Could the lack of report from the ice detectors be due to a different type of ice being present, than they were designed / able to detect? Being only a GA pilot, I have no idea how the ice detectors on an A330 work - any pointers?

After the logged AoA sensor #1 value gets "unstuck" from the 2.1deg value (thanks takata for that), at around 02:10:52, note that its value then gets "stuck" again for a while, but this time at a much higher value (> 40 deg) at around 02:11:45, when the other two AoA sensor values dip below 40 deg (page 107 English version).

My overall view is that there is a general "stickiness" to that sensor, and that's telling us something - whether it's affected by ice (of a type / amount / location not picked up by the ice sensors), or is a faulty sensor (lack of effective heating?), or ... ?

As I said, I hope the AoA #1 output wasn't being used due to earlier discrepancies, but I haven't seen that confirmed anywhere (I'd be very happy to be corrected on that!).
Diagnostic is offline  
Old 8th Aug 2011, 23:34
  #1779 (permalink)  
 
Join Date: Jul 2009
Location: UK
Posts: 134
Likes: 0
Received 0 Likes on 0 Posts
Has it been confirmed what AoA sensors were fitted?

Pretty likely the Thales C16291AA given the choice of Pitot tube, if so one could have been suffering an extreme variation of the known fault at low temperature:-

EASA Airworthiness Directive

Note that the use of 3 sensors means that a single failure can be accommodated - but surely there should eventually be an ACARS maintenance warning to get the probe looked at?

Last edited by sensor_validation; 9th Aug 2011 at 09:54.
sensor_validation is offline  
Old 8th Aug 2011, 23:41
  #1780 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
There was a direct relationship between the rolling and pitching moments and that is best illustrated by superimposing the rolling data over that of pitching. Also remember that at all times the yaw damper continued to work and the rudder was being deflected up to the maximum of +/-7.9 allowed by the RTLU. Worth checking is the Normal Acceleration [g] to which has been added a smoothed line (red), and observe the 'g' recorded during the period of the THS movement.

The graphic below is a rejigged composite of those presented by the BEA, with the exception that the topmost section shows CAS as calculated from BEA data by HazelNuts39.

The sections covered by a pale yellow background are related to the period in which the THS moved from 3NU to 13.6NU, and the sections with a pale green background draw attention to the effect of a reduction in thrust.

During periods of prolonged right rolling/banking, the nose dropped in unison and the heading changed.

If the situation had been understood by no later than 2:12:10, there was a very good chance that the ND bank then taking place could have been converted into sufficient airspeed to allow a coordinated recovery. It didn't happen, as continued NU inputs and application of TOGA thrust saw to that.

Last edited by mm43; 9th Aug 2011 at 02:20. Reason: grammar
mm43 is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.