B-738 Crash in Russia Rostov-on-Don
Join Date: Apr 2008
Location: Australia
Posts: 669
Likes: 0
Received 0 Likes
on
0 Posts
A few posters here seem perplexed at how a pilot wouldn't "instinctively", or "reflexively" reverse his pitch trim inputs after the aircraft rapidly entered the dive.
These posters perhaps don't understand human behaviour very well. Put a human under enough stress, and they won't make the same responses they would have when calm.
Here is an example. In 2004, a pilot of an Aerocommander AC500S engaged the autopilot shortly after reaching top of climb. Unfortunately, the design of the human/machine interface in the case of this autopilot was very poor.
The autopilot immediately pitched nose down, and proceeded to wind on full (or near full) nose-down trim. The aircraft, in a steep dive, exceeded limit airspeeds and disintegrated in flight.
ATSB investigation report here: https://www.atsb.gov.au/media/24274/...400610_001.pdf
It appears that the pilot was startled by the immediate and violent pitch down, and responded by pulling on the CC - rather than by disengaging the autopilot.
These posters perhaps don't understand human behaviour very well. Put a human under enough stress, and they won't make the same responses they would have when calm.
Here is an example. In 2004, a pilot of an Aerocommander AC500S engaged the autopilot shortly after reaching top of climb. Unfortunately, the design of the human/machine interface in the case of this autopilot was very poor.
The autopilot immediately pitched nose down, and proceeded to wind on full (or near full) nose-down trim. The aircraft, in a steep dive, exceeded limit airspeeds and disintegrated in flight.
ATSB investigation report here: https://www.atsb.gov.au/media/24274/...400610_001.pdf
It appears that the pilot was startled by the immediate and violent pitch down, and responded by pulling on the CC - rather than by disengaging the autopilot.
Add to it the chronique inability of airline pilots to scan instruments coupled with tiredness, somatographic illusion and being behind the aircraft due to a lack of training as was mentioned in previous post.
I think we will see many of such accidents in the future. With the lack of training airline pilots are becoming like a sheep with no survival instincts. They cannot think one step ahead of the aircraft. Everything is a surprise to them.
And doing the 2 previous goarounds fine, while on autopilot and the last one badly because of manual flying... what is so difficult to understand it went wrong? High energy goaround keeps killing pilots with lack of instrument scanning ability, period.
I think we will see many of such accidents in the future. With the lack of training airline pilots are becoming like a sheep with no survival instincts. They cannot think one step ahead of the aircraft. Everything is a surprise to them.
And doing the 2 previous goarounds fine, while on autopilot and the last one badly because of manual flying... what is so difficult to understand it went wrong? High energy goaround keeps killing pilots with lack of instrument scanning ability, period.
Join Date: Sep 2005
Location: Moscow, Russia
Posts: 1,011
Likes: 0
Received 0 Likes
on
0 Posts
This posted video was made by Chief Pilot inspector of S7/Globus Denis Okan. He was specially tried to repeat accident condition. He has large post in his LJ with throughout explanation about this simulation.
Accident scenario - -= Fly Safe! =-
Accident scenario - -= Fly Safe! =-
Join Date: Aug 2008
Location: Betwixt and between
Posts: 666
Likes: 0
Received 0 Likes
on
0 Posts
Chuks said:
However, this I agree with 100% This is a concern of mine in this event.
Chukss also said:
If you look at my posts, I am with the notion that a crew who already performed a go-around and were expecting to make another go-around and then flew what appears to be a stable second go-around with all the available parameters consistent with such, then suddenly end up diving at 13,500 FPM in 7 seconds from a 1350 FPM climb cannot possibly be explained away with just crew error.
However, I can understand how, once trimmed fully nose down, if the speed is allowed to builds sufficiently, then even full aft elevator may be enough to recover the aircraft. With 4 units of trim, I could never have imagined it but I can accept it. With zero units and the flaps ripped of I can easily believe it.
But, what is most puzzling is why the half way down the descent, the flight path steepened still further. Any attempt to pull back on the stick would have resulted in, at worst, a pitch up, however slight and whatever the trim. Perhaps there is a point at which the elevator becomes stalled and looses all effectiveness?
But as Chuks ponders, if there was nothing wrong with the aircraft, why not simply trim nose up. It is inconceivable how this was not the first and initial reaction of even the most fatigued, surprised, shock, scared and startled crew member, never mind crew members!
BTW, any bank angle can easily be explained away. At high speed, even the slightest and gentlest input will result in an instant response and roll - even with no intention to bank the aircraft it would be unlikely that a crew trying to recover won't command some roll at such speeds. The aileron feel system uses only springs and does not replicate dynamic aerodynamic loads in the same way as the elevator system.
Strange things do happen; there have been other crashes that require reading the final report to understand how they could have happened, when that was because of very strange control inputs on the part of the crew. (That Colgan crash near Buffalo is probably the worst example of that, when even the accident investigators had no real answer to why the crew flew the airplane as they did, when their strange control inputs caused the crash.) Here we have an equally strange scenario being presented in a veiled manner, out of fairness to the official inquiry (?). Okay, if you say so!
Chukss also said:
It's completely unrealistic for OKC to tell us of a scenario where he is pulling with all his might against the trim without bothering to re-trim using switches that are right there on the yoke. Who would do that, especially given that they would have been using those same switches shortly before to re-trim the aircraft? It is not as if the PF is going to suddenly forget to use the same secondary flight control, pitch trim, that he already had been using, is it? This is not, at first glance, a realistic scenario, so that we need to be told more about how it's supposed to have come about.
However, I can understand how, once trimmed fully nose down, if the speed is allowed to builds sufficiently, then even full aft elevator may be enough to recover the aircraft. With 4 units of trim, I could never have imagined it but I can accept it. With zero units and the flaps ripped of I can easily believe it.
But, what is most puzzling is why the half way down the descent, the flight path steepened still further. Any attempt to pull back on the stick would have resulted in, at worst, a pitch up, however slight and whatever the trim. Perhaps there is a point at which the elevator becomes stalled and looses all effectiveness?
But as Chuks ponders, if there was nothing wrong with the aircraft, why not simply trim nose up. It is inconceivable how this was not the first and initial reaction of even the most fatigued, surprised, shock, scared and startled crew member, never mind crew members!
BTW, any bank angle can easily be explained away. At high speed, even the slightest and gentlest input will result in an instant response and roll - even with no intention to bank the aircraft it would be unlikely that a crew trying to recover won't command some roll at such speeds. The aileron feel system uses only springs and does not replicate dynamic aerodynamic loads in the same way as the elevator system.
Join Date: Aug 2000
Posts: 1,501
Likes: 0
Received 0 Likes
on
0 Posts
They added a lot of power during the goaround. It is normal and expected. Than the nose goes up a lot during the power application. So you have to push a lot to keep the nose at 15 degrees or to get it down to this 15 degrees, in case you slept through it
A normal two engine go around in the NG is not a violent, lots of power event. There is no need to push a lot or trim forward go get the nose to 15 degrees. It pretty much parks there if everything works OK.
If, for whatever reason they decided to use full thrust during go around in a light NG, it could get interesting really fast.
This was a single channel automatic approach with a manual go around. The autopilot disconnected just at it should do. FZ SOP.
OKC is a very experienced pilot. Stop making him look like an idiot because he did a typo. He made a post about trying to replicate this crash, but unless I am mistaken, the video posted is not his.
But, what is most puzzling is why the half way down the descent, the flight path steepened still further. Any attempt to pull back on the stick would have resulted in, at worst, a pitch up, however slight and whatever the trim. Perhaps there is a point at which the elevator becomes stalled and looses all effectiveness
Mostly Agree Mana, although the 737 at full TOGA thrust, especially at light weight, does pitch up and can be a handful. That's why there are two TOGA modes with the AT engaged. A single push of the TOGA button results in pitch and power approximating a 1500 fpm climb while a two push TOGA gives full TOGA thrust. There is a reason for that feature. You sort of said that, but kind of contradicted it in your admonishment of a post as well.
Stabilizer is the boss? Spot on!
Sciolistes,
You inquire why the path steepened but really you answered your own question at the same time. As the airspeed increased the "crossover" effect of the dominant stabilizer authority probably continued to increase thus driving the nose down more and more as AS increased. Also, I don't find it inconceivable that a very fatigued crew under certain circumstances could react this way during a TOGA.
Stabilizer is the boss? Spot on!
Sciolistes,
You inquire why the path steepened but really you answered your own question at the same time. As the airspeed increased the "crossover" effect of the dominant stabilizer authority probably continued to increase thus driving the nose down more and more as AS increased. Also, I don't find it inconceivable that a very fatigued crew under certain circumstances could react this way during a TOGA.
Join Date: Aug 2000
Posts: 1,501
Likes: 0
Received 0 Likes
on
0 Posts
I said in a normal go around. Again, there os nothing violent about a normal go around in the NG.
A full thrust go around is not a normal procedure in a light NG.
I have said long time ago that full thrust combined with a "surprise" level off could go wrong very fast.
If you trim way forward, don't be surprised if the nose suddenly points down. It is designed to do that.
What surprised me in the video was how little time it took from level, pitch down to crash.
A full thrust go around is not a normal procedure in a light NG.
I have said long time ago that full thrust combined with a "surprise" level off could go wrong very fast.
If you trim way forward, don't be surprised if the nose suddenly points down. It is designed to do that.
What surprised me in the video was how little time it took from level, pitch down to crash.
This was a single channel automatic approach with a manual go around. The autopilot disconnected just at it should do. FZ SOP.
1. Was the autothrottle engaged?
2. If thrust was being managed manually was the autothrottle ARM mode used?
3. Was a normal go-around flown or was it a Windshear escape manoeuvre?
There are many unanswered questions of "what" before we get into "why"
However, as many have already suggested, fatigued people may do strange things or end up behind the aircraft. I'm sure we have all done daft things in more benign conditions and been saved by altitude, reacquiring situational awareness, or the person in the other seat.
I have also operated while fatigued (not in an aircraft and during time in the military) and the ability to make basic mistakes in that condition is quite shocking.
On a regulatory front, I hope this will at the very least bring to the end the farce that the head of the Emirati regulatory body also happens to own EK and FZ...
Join Date: Dec 2006
Location: Florida and wherever my laptop is
Posts: 1,350
Likes: 0
Received 0 Likes
on
0 Posts
Several studies have pointed out that fatigue (often called drowsiness) causes as many errors as being under the influence of alcohol. This should be pointed out to the roster managers in the airlines that appear to be using maximum hours limits as normal rostering hours.
If airlines will not allow their pilots to have a beer before duty, then they should show as much concern to flying tired which is now being shown to be as likely to cause errors.
Perhaps gate checks should include the crew's previous months rostering.
In this case the crew also were put into a difficult Human Factors position one that was pointed out in 1908 by Yerkes and Dodson http://pe-arousal.********.com/2011/...-u-theory.html (mentioned earlier in the thread) they went from relatively high arousal on the approach to doing a routine go around albeit for just 2 minutes no doubt relaxing back a little to the arousal/boredom of the previous 2 hours holding, then all of a sudden everything starts to go pearshaped and they are at the other end of the arousal spectrum - in HF terms that is going direct from one error prone state to another. Fatigue and over/under arousal are almost guaranteed to lead to errors.
If airlines will not allow their pilots to have a beer before duty, then they should show as much concern to flying tired which is now being shown to be as likely to cause errors.
Perhaps gate checks should include the crew's previous months rostering.
In this case the crew also were put into a difficult Human Factors position one that was pointed out in 1908 by Yerkes and Dodson http://pe-arousal.********.com/2011/...-u-theory.html (mentioned earlier in the thread) they went from relatively high arousal on the approach to doing a routine go around albeit for just 2 minutes no doubt relaxing back a little to the arousal/boredom of the previous 2 hours holding, then all of a sudden everything starts to go pearshaped and they are at the other end of the arousal spectrum - in HF terms that is going direct from one error prone state to another. Fatigue and over/under arousal are almost guaranteed to lead to errors.
Join Date: Jan 2000
Location: FL410
Posts: 860
Likes: 0
Received 0 Likes
on
0 Posts
1. Was the autothrottle engaged?
2. If thrust was being managed manually was the autothrottle ARM mode used?
3. Was a normal go-around flown or was it a Windshear escape manoeuvre?
2. If thrust was being managed manually was the autothrottle ARM mode used?
3. Was a normal go-around flown or was it a Windshear escape manoeuvre?
2. Once TOGA is pressed, AT move up, if AT had been disconnected prior on approach, AT will not react further unless re-armed by crew during re-engagement sequence, usually when aircraft is being levelled out in trim.
3. It does not matter in this case, unless you are attempting to establish if crew (should have) disconnected AT as part of flown procedure.
Here is a thought:
On an ACQ manoeuvre, MCP speed window opens, if a low level altitude acquire happened then speed will be actual rather then associated flap speed, thus AT when selected will be reducing power with open speed window to establish commanded speed rather then required speed.
Add to this the possible FDs commanding the lower altitude which has been bust (as not set the higher FL080 in MCP) and FDs wanting to fly downwards (pitch forward), you need not think too long about what can happen next, when ballooning trying to re-ACQ the overshot altitude.
Another problem could be a vertical mode selection on MCP by crew: if the procedure is coded to have lower level off but MCP Altitude is set correctly at FL080, when Flaps Up, an MCP selection of VNAV may result in a descend under certain conditions when selected...
Could either of these things have happened?
1/ two taps of A/T instead of one? Yip.
2/ A whole handful of thrust if A/T not engaged? Yip.
So a simple error ( either 1 or 2) and you now have a lite NG with full power climbing into an increasing headwind at 4am with a crew who has been airborne for six hours, one of whom allegedly resigned due fatiguing rosters.
It is not difficult to imagine how the above situation is only a few moments of brain-fart away from tragedy.
So if this turns out to be the case, what should be done?
Regulate to ensure our Airline crews aren't rostered to the point of fatigue.
They should be the opposite. Mentally fit, well rested, well trained.
1/ two taps of A/T instead of one? Yip.
2/ A whole handful of thrust if A/T not engaged? Yip.
So a simple error ( either 1 or 2) and you now have a lite NG with full power climbing into an increasing headwind at 4am with a crew who has been airborne for six hours, one of whom allegedly resigned due fatiguing rosters.
It is not difficult to imagine how the above situation is only a few moments of brain-fart away from tragedy.
So if this turns out to be the case, what should be done?
Regulate to ensure our Airline crews aren't rostered to the point of fatigue.
They should be the opposite. Mentally fit, well rested, well trained.
Join Date: Aug 2008
Location: Betwixt and between
Posts: 666
Likes: 0
Received 0 Likes
on
0 Posts
737er,
No, it is not inconceivable. But this appears to have happened at the end of a what, by profile and speed, appeared to be an adequately executed missed approach. Long after the TOGA button was pressed and after the acceleration phase was complete or near the end of that phase. So it is inconceivable that they managed to control the aircraft at the critical point and end up with such an extreme and sudden nose over when the aircraft was what appears to be, completely stable.
Also, I don't find it inconceivable that a very fatigued crew under certain circumstances could react this way during a TOGA.
Join Date: Jan 2015
Location: Centre of Universe
Posts: 371
Likes: 0
Received 0 Likes
on
0 Posts
Several studies have pointed out that fatigue (often called drowsiness) causes as many errors as being under the influence of alcohol. This should be pointed out to the roster managers in the airlines that appear to be using maximum hours limits as normal rostering hours.
Limits - yes regretfully the world we live in these days. No use pointing at Roster Managers. NAA's, EASA, IACA etc
Join Date: Mar 2014
Location: Brussels
Posts: 131
Likes: 0
Received 0 Likes
on
0 Posts
As to the use of the trim, if you have just been using the trim switches, yielding some unpleasant result, how in the world are you suddenly going to forget all about the use of the trim switches, as in pushing them the other way from the way you just did push them? That would be a brain-fart of cosmic proportions.
Well, that's a Middle Marker tone, I think.
Well, that's a Middle Marker tone, I think.
As to the question of the 'noise' on the video, the cryptic poster was referring to the trim-wheel wizzing for quite some considerable time. This is not a noise you are likely to miss, to comment on, and to react to, if it happened of its own accord. So perhaps we can rule out mechanical failure.
ST
Join Date: Jan 2008
Location: Hotel Sheets, Downtown Plunketville
Age: 76
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
The 738 is an aircraft designed for automatics. It has STS which basically is a speed stability
augmentation system designed to improve flight characteristics during during operations with low GW, aft CG and high thrust when a/p is not engaged. STS monitors the stab input and thrust lever positions, air speed and vertical speed and trims the stab with the a/p stab trim. The STS operates counter to pilot inputs.
Somewhere therein may lay the answer to this puzzle.
augmentation system designed to improve flight characteristics during during operations with low GW, aft CG and high thrust when a/p is not engaged. STS monitors the stab input and thrust lever positions, air speed and vertical speed and trims the stab with the a/p stab trim. The STS operates counter to pilot inputs.
Somewhere therein may lay the answer to this puzzle.
There is no "trim in motion" sound on the 737, merely the generally noticeable sound of the trim wheel spinning. The "sound that explains everything" is the 5 seconds of the trim wheel running. When I am next working I will have a look how much of the trim range 5 seconds of trim at high rate covers, but I would guess it is not far off half the total trim range.
Most who fly the 737 will subconsciously be aware of any trim movement, and look to see what caused it, especially after the Turkish accident, which I suspect almost every 737 driver has relived in the sim. In that case, the biggest clue (apart from the speed and the speed trend dropping through the floor) was the trim wheel running considerably longer than it would in any normal condition.
Most who fly the 737 will subconsciously be aware of any trim movement, and look to see what caused it, especially after the Turkish accident, which I suspect almost every 737 driver has relived in the sim. In that case, the biggest clue (apart from the speed and the speed trend dropping through the floor) was the trim wheel running considerably longer than it would in any normal condition.
Join Date: Aug 2008
Location: Betwixt and between
Posts: 666
Likes: 0
Received 0 Likes
on
0 Posts
Chucks,
The trimmer in that video was running for 12 seconds! But the video is misleading as they had not completed the flight profile that is evident in the FR24 data and the flaps were still extended - which is why it rolled so much too. Disregard.
heard earlier, but it does not run for very long. The Middle Marker tone is the only one sounding over that whole period.
Limits - yes regretfully the world we live in these days. No use pointing at Roster Managers. NAA's, EASA, IACA etc
The "sound that explains everything" is the 5 seconds of the trim wheel running. When I am next working I will have a look how much of the trim range 5 seconds of trim at high rate covers, but I would guess it is not far off half the total trim range.
I also get a "Stabiliser Motion" call from "Patricia" if the trim runs continuously more than 1 degree.