Go Back  PPRuNe Forums > Flight Deck Forums > Rumours & News
Reload this Page >

Ethiopian airliner down in Africa

Rumours & News Reporting Points that may affect our jobs or lives as professional pilots. Also, items that may be of interest to professional pilots.

Ethiopian airliner down in Africa

Old 12th Apr 2019, 22:35
  #3921 (permalink)  
fdr
 
Join Date: Jun 2001
Location: 3rd Rock, #29B
Posts: 798
Comments re thrust on ET302 seem to suggest that from an arm chair the crew can be criticised for keeping it at high thrust.

The BAC FCTM states to attempt to match the out of trim speed in order to reduce load on the trim system. The aircraft is trimmed to an extremely high warp factor (aoa) and the inability to trim a correction arises due to the attendant forces applied by the off trim of the stab, and the corrective elevator input.

Reducing power as suggested does not achieve the matching of speed necessary to unload the stab, yet the comments suggest that the crew erred by keeping power on.... really? The fastest way to restore control, with the least altitude loss comes in this case by having high thrust applied.

Now, in the first instance of a problem with trim, keeping speed in check sounds like a nice principle, but that presumes that the trim has not run away fully, in which case, the control problem is exacerbated by the speed error between current and trimmed speed (aoa).

These guys had a substantial problem to fix in a time critical event, and the comments critical of their technique show the general malaise and lack of knowledge that exists in this industry. Hard to call it a profession. If criticising their actions at least get the physics right.
fdr is offline  
Old 12th Apr 2019, 22:45
  #3922 (permalink)  
 
Join Date: May 2010
Location: Boston
Age: 69
Posts: 443
Originally Posted by VicMel View Post
...
...
Whatever damping there may be in the AoA sensor it has to be fairly light otherwise there would be a significant delay on the AoA signal and obviously the pilot does not want to be told he is approaching a stall condition after he has got there! More seriously, if the counterweight without a vane can really get from 35.7 deg to 74.5 deg in 0.75 of a second, then there is very little damping. I do not think it is feasible that the counterweight can then stop, nearly instantaneously, onto the 74.5 deg position, without any overshoot at all, or bouncing off of any end stop that is set at 74.5 deg. Naah, you need software to do that sort of screw up.
...
...
Secondly, good point, Iíve no idea why the L AoA Heat changed state. Iím puzzled as to why the state was On at the beginning, when the outside temperature was 16deg C. However, Iím more inclined to believe that it was spurious than I am to believe there was anything other than a single underlying failure which affected all 3 of the flights.

Vic
The damping is set for significantly higher forces than the weight itself which is just to counterbalance the vane.
The end stop could well be a 'soft' stop where things get wedged, the ticks downward from ~.5g spikes suggest just that.

If you compare the reported left AoA position to vertical g at the end it is a close to perfect match.

The AoA vane heat is likely on by default, it does get a bit colder at higher altitudes. Also there was a reason it was included in the prelim report, another 1000+ (I may be off here) parameters were not.
MurphyWasRight is offline  
Old 12th Apr 2019, 23:57
  #3923 (permalink)  
 
Join Date: May 2016
Location: Nantes
Posts: 63
Originally Posted by Cows getting bigger View Post
I think there is another HF issue (already commented on some time back) in that it is utterly counter-intuitive to reduce thrust when the nose is pitching down. It is like leaning to right when you want to turn your bicycle to the left.
Reducing speed would have been also against what Boeing recommends in case of a severe AND mis-trim, which is increasing speed (not above VMO, though) in the hope of reducing elevator loads
deltafox44 is offline  
Old 13th Apr 2019, 00:17
  #3924 (permalink)  
 
Join Date: May 2016
Location: Nantes
Posts: 63
Originally Posted by VicMel View Post
Firstly, the AoA values are all driven by the correction algorithm, as I have theorised above, mainly the one using Pitch Rate as a driver.
I don't know if AoA value shown on the graph was value from the vane or value corrected by algorithms. I suppose both are recorded. Anyway, pitch rate correction cannot drive corrected value from +75į to -60į !
deltafox44 is offline  
Old 13th Apr 2019, 00:29
  #3925 (permalink)  
 
Join Date: Feb 2019
Location: shiny side up
Posts: 431
Firstly, the AoA values are all driven by the correction algorithm, as I have theorised above, mainly the one using Pitch Rate as a driver.
Thank you! Finally, someone else who understands that the output from the AoA vane is subject to numerous algorithms, depending on conditions. To completely base pending stall conditions, and an automated response based on the AoA vane measurement is foolhardy.

First off, the algorithm corrects between the location of the AoA vane, the CG, and wing location angles of attack. The algorithm then corrects for airspeed and sideslip. What is interesting is the AoA measurement becomes less accurate in a climb, especially in a climbing turn (per Boeing) as the winds around the fuselage are deflected, shielded, or otherwise disrupted.

The best solution would be to mount the AoA vane on the wing at midspan, but the disruption from the engine, and moving leading edge prevents this.

The engines are mounted higher on the wing, and further forward, yet the AoA location remained unchanged. Another indication is the green band, on the 738, it is from 1 to 6, yet on the 737-8, it is 3 to 8.....

As noted, originally, the software and modelling people came up with a 0.6 correction on the trim. In testing, that turned into 2.5 degrees. This tells me that the assumptions that created the algorithm are not correct.
Smythe is offline  
Old 13th Apr 2019, 01:16
  #3926 (permalink)  
 
Join Date: Nov 2018
Location: madrid
Posts: 47
AOA corrects airspeed and then airspeed corrects AOA?
ecto1 is offline  
Old 13th Apr 2019, 01:26
  #3927 (permalink)  
Psychophysiological entity
 
Join Date: Jun 2001
Location: Tweet Rob_Benham Famous author. Well, slightly famous.
Age: 81
Posts: 4,898
Thanks for the replies.

The double blips. I begged the question some 1,000 posts ago about the possibility of the blips causing alarm because of sudden g forces, despite the graphs not showing any such detail. Strangely, the higher accelerative fluctuations come later - I imagine distressing belt-tugging forces.

The consensus back then dismissed the possibility. So, looking again for why the trim wasn't continued. The extract above is again thought provoking; giving an insight into the psychology. But I'm still not convinced. The g force noise does go up during that period, but not in any recognisable sustained pattern. The column inputs are very well mirrored in g forces shortly after.

05:41:35-40 -ish shows a sustained average positive g despite coming from a series of no doubt powerful tugs. Since it seems to be real, and showing in a definite improved pitch attitude, one wonders why this could not be sustained.

All in all, I still have great sympathy for the pilots, which is no doubt strongly biased by my short flight with the stick shaker going. My highly competent FO pointed in turn at the six speed and attitude dials, but even then half a lifetime of indoctrination was relentlessly tugging at my mind, and that's with nothing else being wrong.
Loose rivets is offline  
Old 13th Apr 2019, 02:02
  #3928 (permalink)  
 
Join Date: Feb 2019
Location: shiny side up
Posts: 431
Originally Posted by ecto1 View Post
AOA corrects airspeed and then airspeed corrects AOA?
No, they do not correct each other.
The stall algorithm uses a varied number of inputs to determine if the ac is getting into a stall, most of which are algorithm based due to latency.
Latency of all of the calculations is another issue.
While the critical angle of attack doesnt change, the various factors involves such as weight, configuration, CG, and winds, change the airspeed at which an ac will stall.

Now you have unreliable airspeed indicators, (since the pitot tubes do not move into the winds, they are shielded on turns, angle of attack and crosswinds) unreliable AoA measurements, and a varying configuration feeding an automated trim correction.
Smythe is offline  
Old 13th Apr 2019, 04:17
  #3929 (permalink)  
 
Join Date: Jan 2013
Location: Seattle Area
Posts: 237
Originally Posted by Smythe View Post
No, they do not correct each other.
.
I'm quite sure that's not fully correct. There is a correction to same side static pressure based on AOA, which affects both airspeed and altitude on the same side. Airspeed is not used to modify the sensed or displayed AOA as far as I know.
Dave Therhino is offline  
Old 13th Apr 2019, 06:04
  #3930 (permalink)  
 
Join Date: Sep 2018
Location: Laredo, TX
Posts: 123
Originally Posted by fdr View Post
Comments re thrust on ET302 seem to suggest that from an arm chair the crew can be criticised for keeping it at high thrust.

The BAC FCTM states to attempt to match the out of trim speed in order to reduce load on the trim system. The aircraft is trimmed to an extremely high warp factor (aoa) and the inability to trim a correction arises due to the attendant forces applied by the off trim of the stab, and the corrective elevator input.

Reducing power as suggested does not achieve the matching of speed necessary to unload the stab, yet the comments suggest that the crew erred by keeping power on.... really? The fastest way to restore control, with the least altitude loss comes in this case by having high thrust applied.

Now, in the first instance of a problem with trim, keeping speed in check sounds like a nice principle, but that presumes that the trim has not run away fully, in which case, the control problem is exacerbated by the speed error between current and trimmed speed (aoa).

These guys had a substantial problem to fix in a time critical event, and the comments critical of their technique show the general malaise and lack of knowledge that exists in this industry. Hard to call it a profession. If criticising their actions at least get the physics right.
The crew did not ďkeepĒ the thrust high. They ignored a control available to them and ignored its effects upon the aircraft. There was no cognitive thought process that said letís go past Vmo because that will be better for us.
jimtx is offline  
Old 13th Apr 2019, 09:09
  #3931 (permalink)  
 
Join Date: Dec 2002
Location: UK
Posts: 2,076
VicMel, an interesting analysis #3958. Together with that of Smythe … now confused.

(If) Vane-sensed AoA requires correction; is this correction in the vane unit or elsewhere. Peter Lemme argues that errors originate from the vane unit as AoA output is ‘sent’ and used by many separate boxes.
Alternatively a separate (remote) calculation would be more logical with the need of other factors, e.g. embedded in ADC for Mach; which box (function) is not so important if the corrected AoA is available on a digital bus to all other computational boxes. This would also negate the view of a vane-origin error.

Airspeed pressure-error corrections depend on the corrected AoA; differences between ADC computations are alerted (Speed, Alt, Disagree). This should not change the AoA.

AoA is use to position the low speed awareness symbol on the airspeed scale, but this too uses output AoA, it should not correct - change either AoA or airspeed.

Thus what correction is being discussed; where is the computation done, and how is the output accessed by many other functions requiring AoA - MCAS one amongst many.

Re vane damping; an earlier suggestion was that the vane slipped position on the shaft; fortuitously catching at different times and values in each accident. A slipping vane/shaft, when snagged would then act as a single unit involving aerodynamic and mass damping - but incorrect AoA.
Although unlikely, Occam might like the simple mechanical approach.

Alternatively, the digital views might follow Moore-Murphy, where with increasing complexity and add-ons, there is even greater risk of error, the need crosschecking and multiple sources would be paramount. Somewhat lacking in this system.




safetypee is offline  
Old 13th Apr 2019, 09:33
  #3932 (permalink)  
 
Join Date: Jan 2019
Location: Earth
Posts: 26
Originally Posted by Capn Bloggs View Post
ex-USAF Major General. Says it all. Blame the pilots when in reality they were flying a deathtrap. What are we, testpilots, finely honed, ready to pounce into action when disaster strikes?
Seekingalpha is an investment-related blog, anybody can write whatever they like there. There are two kinds of people on that site: those talking up stock they own, and those talking down stock they sold short. Utterly unreliable as a source of investment info, let alone anything else.
delarue is offline  
Old 13th Apr 2019, 09:44
  #3933 (permalink)  
 
Join Date: Mar 2002
Location: Seat 0A
Posts: 8,031
Delarue, the thing that annoys me is he sounds credible, and that the riff raff probably believe him.
Capn Bloggs is offline  
Old 13th Apr 2019, 09:52
  #3934 (permalink)  
 
Join Date: Feb 2013
Location: 60 north
Age: 56
Posts: 20
Auto Throttle

At one point the FO selected 238kts in the speed window.
I think that was the only time I can see they try to control speed. and a good speed indeed.
My goto speed is 230kts non the 737-800 at this stage as this gives me any options and safe margins.
BUT
The AT must have disconnected at one point as the IAS became different.
BluSdUp is offline  
Old 13th Apr 2019, 10:09
  #3935 (permalink)  
 
Join Date: Jul 2014
Location: Harbour Master Place
Posts: 661
Originally Posted by BluSdUp View Post
The AT must have disconnected at one point as the IAS became different.
They probably had the AT disconnect annunciator flashing away until the end.

When I see that AT disconnect flashing for more than a few seconds, I know the person I am sitting next to is overloaded and has lost SA. It's interesting experiment to see how long before they attend to it, it can be quite some time. Occasionally they won't even notice I've reached up and pushed the annunciator to cancel it.
CurtainTwitcher is offline  
Old 13th Apr 2019, 11:45
  #3936 (permalink)  
 
Join Date: Jun 2001
Location: UTC +8
Posts: 2,624
CurtainTwitcher;"They probably had the AT disconnect annunciator flashing away until the end."

"When I see that AT disconnect flashing for more than a few seconds, I know the person I am sitting next to is overloaded and has lost SA. It's interesting experiment to see how long before they attend to it, it can be quite some time. Occasionally they won't even notice I've reached up and pushed the annunciator to cancel it."

...Except that it's hard not to hear the annoying clacker horn as they zoomed past 340 kts (VMO) 3 minutes after lift off, and past 458 kts two minutes after that. It's apparent that these guys were totally engrossed with pitch control ignoring also the increasing noise of rushing air from the slipstream, ignoring also in day VMC at low altitude the fast moving terrain below their feet.

...
GlueBall is offline  
Old 13th Apr 2019, 12:52
  #3937 (permalink)  
 
Join Date: May 2016
Location: Nantes
Posts: 63
Originally Posted by Smythe View Post
The best solution would be to mount the AoA vane on the wing at midspan, but the disruption from the engine, and moving leading edge prevents this.
No, the wing would be the worst place to measure AoA, since airflow perturbation is the highest there. The only place to get an good AoA is on a long stick ahead of the nose. Even there it would be correct only in steady flight in still air...

Last edited by deltafox44; 13th Apr 2019 at 13:07.
deltafox44 is offline  
Old 13th Apr 2019, 14:28
  #3938 (permalink)  
 
Join Date: Jun 2009
Location: Europe
Age: 51
Posts: 11
Originally Posted by L39 Guy View Post
Based upon my 36 years/26,000 hours of flying, provided that one does not "jerk" the aircraft into the air, if one rotates the aircraft nose at a normal rate to the take-off attitude, roughly 15 degrees in the B737 at 3 degrees per second, the aircraft go flying when the wings are ready to let it go flying, i.e. when the have created enough lift. And, with both engines running, the aircraft will accelerate. Because of this characteristic, this will cover up a lot of mistakes such as wrong flap setting, wrong power setting, wrong C of G (although that would be more an issue of control column forces to rotate as the stab trim setting is based upon the C of G). This would mean the aircraft would take more runway if heavier, lower flaps setting (1 instead of 5 for example), lower power setting and less runway with the opposite conditions.

Regardless of any of this, surely by 400 ft there would be communication between the two pilots about the indicated airspeed between the three airspeed indicators in the flight deck, particularly since the Captain was an 8000+ hour pilot; the FO was probably shell shocked.

But either way - unreliable airspeed or a bona fide stall - why on earth would the Captain call "Command", i.e autopilot engagement at 400 ft? You do not engage an autopilot when one is in a stall, you do not engage an autopilot with an unreliable airspeed.

To me, this points to an experience, training and attitude problem the world over in modern airline flying. Pilots no longer have the basic flying skills to fly an airplane anymore without autopilot, autothrottle, flight director and, heavens forbid, GPS/RNAV! Where I fly (Canada) we call these people "Children of the Magenta Line". I don't know if other countries use this term or not but what it means is all they know is how to fly the magenta line on the nav display as well as the magenta pointers on the primary flight display.

I know in other parts of the world, hand flying an aircraft is not only discouraged its against SOP's and is subject to the "FDR Police". It's all fine and dandy when things are going well but when the crap hits the fan and one actually has to revert to basic flying skills they are not there, either because they never were there in the first place or, if they were, they have atrophied because they haven't been used in years.

While the MAX incident/accidents have brought this home tragically - trying to use the autopilot in a unreliable airspeed situation or a true stall (take your pick), being unable to trim the aircraft with the electric trim (continuous trim rather than short bursts), being unable to manually trim and fly the aircraft, being unable to manage the airspeed and not going the speed of heat, or questionable airmanship decisions such as continuing to destination with unreliable airspeed, and so on - there have been scores of incidents such as the Korean 777 in SFO wherein the crew could not fly a visual approach on a clear day, etc. Only by pure luck or the incredible survivability of the aircraft that no one was killed in the crash itself. And there are scores of other examples of incidents that could have easily become fatal accidents, just go to avherald.com to see for yourself.

The entire industry - ICAO, IATA, the individual airlines, the individual CAA's, the pilot unions, aircraft manufacturers, etc - need to do some serious navel gazing to get the level of pilot proficiency and training back to the point where paying customers can count on the pilot to be the last line of defense when the unexpected happens such as a double engine failure (US Air). Technology is great but it has its limitations and at the end of the day, trained and competent pilots are still needed when the unanticipated events happen.

Editorial over.
They even tried to engage the autopilot near the end at 05:43:15, hoping the automation would fix their problems!
It seems the reason they switched the cutoff-switches back was so they could reengange the autopilot.

This is certainly a severe case of "the children of the magenta line".

Last edited by Brosa; 13th Apr 2019 at 17:56.
Brosa is offline  
Old 13th Apr 2019, 15:34
  #3939 (permalink)  
 
Join Date: Mar 2019
Location: Vantaa, Finland
Posts: 0
It is some while ago I had to handle the airfoil basics but I think that upwash field is affected by speed and fuselage angle of attac affect the side mounted vane. So both speed and fuselage AoA may have an effect that might be taken into account when correcting the vane reading.
Aihkio is offline  
Old 13th Apr 2019, 15:55
  #3940 (permalink)  
 
Join Date: Jan 2008
Location: uk
Posts: 847
Originally Posted by MurphyWasRight View Post
That exchange was at/after the electrical trim was re-enabled, I believe the 'manual' refers to electrical trim switches.
Disagree. From the report:

At 05:41:46, the Captain asked the First-Officer if the trim is functional. The First-Officer has replied that the trim was not working and asked if he could try it manually. The Captain told him to try. At 05:41:54, the First-Officer replied that it is not working.
From the FDR traces, no electric trim command (autopilot or manual) is recorded between approx 05:41 and 05:43, therefore the above exchange can only be referring to manual trim with the wheel (unless you are suggesting a simultaneous failure of the trim switches).
infrequentflyer789 is offline  

Thread Tools
Search this Thread

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

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