B-738 Crash in Russia Rostov-on-Don

[PBY]

I understand they wanted to climb to 80. But they were not cleared for it, as far as I know. What if they forgeot and left the GA altitude at 3200, but in their mind they were climbing to 80? Would not that be a big surprise surprise as the autopilot leveld out with hard nose pitch down?

[iggy]

I have been reading the thread since the very beginning and, my two cents, when (the Captain) says "don't worry don't worry" at 3.600' he may be referring to the GA final altitude (3240') overshooting by 250'. It is true that in a high thrust and light weight situation an altitude boost is not that uncommon but maybe this GA altitude boost, if such, may point to a possible AP/trim/FD malfunction that lead to the subsequent pitch down.

What can be hear next is: "don't do that don't do that". Whose voice is it? Is it the same voice as in "don't worry"? If so, we can assume it was the Captain who was flying the GA and then asking the FO not to do something specific when he felt something was off. I'm not a Boeing guy, closest I flew to Boeing was MD80, so here is my question: is there any SOP (Boeing or Fly Dubai) that involves two pilots manipulating the flight controls or surfaces at the same time?

Also, what is the SOP in FD when in GA? Accelerate at final GA altitude or at the standard 1.500' AGL or 3.000' AGL when in noise abatement? If they were planning to level off at final GA altitude (3240') after accelerating at 3000' the flight automation might have acquired the final GA altitude way early (IF it is a function of the v/s during climb, as it usually is, but I am not a Boeing guy), if the FO in that moment cleans up the airplane - without waiting for the Captain command - with a lower than required speed for cleaning, is there any kind of protection in the 737 that could have triggered the nose down moment? Is that why the Captain says: "don't do that?", because the FO retracted the flap below the minimum speed?

Does it make any sense what I'm saying here?

[Craggenmore]

Also, the notion that the pilot accidentally held the trim button down to such an extreme degree is something i find very very hard to believe. It is actually quite ridiculous.

AF 447 more or less held full backstick all the way down from FL350 to impact and the AoA was never less than 35 degrees the report stated.

[Util BUS]

At 6:03 the controller responds to their intention to climb to FL80 with "Roger" although this is non-standard terminology I would have accepted it to me "Approved" in that part of the world.

My experience with the B737 especially regarding operators in the middle east, apart from close to landing, it is unlikely the F/O would have been following through on the controls, if at all.

The two possible maneuvers were a standard go around:

Initiation:
- Push TO/GA, Flaps 15
- Verify go-around attitude
- Verify or adjust thrust as needed
- Positive rate of climb - Gear Up
Above 400 feet RA:
- Verify LNAV or select roll mode
- Verify missed approach altitude set
- Verify route tracking
Acceleration height:
- Retract flaps on schedule
After planned flaps set:
- Select LVL CHG or VNAV
- Verify climb thrust
- Verify altitude capture
- Do after takeoff checklist

Or a windshear escape maneuver:

- Disconnect Autopilot
- Press either TO/GA switch
- Aggressively apply maximum thrust
- Disconnect Autothrottle
- Simultaneously roll wings level and rotate toward an initial pitch attitude of 15 deg
- Retract speedbrakes
- Follow flight director TO/GA guidance (if available)
- Do not change flap or gear configuration until windshear is no longer a factor
- Monitor vertical speed and altitude
- Do not attempt to regain lost airspeed until windshear is no longer a factor

[MrSnuggles]

Anything "knyppel" about this?

https://c2.staticflickr.com/6/5595/14948475956_6f3ded90d4_b.jpg

I don't know anything about the intricacies of a cockpit. This is from a flickr page from a 737NG pilot, as i have understood it.

[GBV]

could flight level 80 be 800 meters? That would be around 3200ft-3500ft which is the altitude they reached during both go-arounds...

[Kulverstukas]

Seems this "CVR transcript" and discussion about stab trim completely put aside previous information from MAK that plane stalled and fell over left wing?

[Daylight Robbery]

I agree, framer. I'm not out, but it's a site for 'professional pilots'. There's some interesting stuff as the awful facts about another accident surface. Would be appreciated if those not flying planes for a living looked and didn't touch.

Regards

[ManaAdaSystem]

The HS on the NG is a very powerful device. If it is grossly out of trim, the elevators are not big enought to get you out of the situation. You have to use the trim as well. Or stop it from happening in the first place.
The problem is this, we are dealing with a system that is not monitored for a runaway condition. Boeing put those big, ugly trim wheels in there, but they live their own life, spinning this way or that way depending on the trim situation. We are used to that, so it takes time to react to a runaway situation. No lady yelling STABILIZER MOTION, STABILIZER MOTION, STABILIZER MOTION on the NG.
Add Speed Trim as framer introduced a few posts ago. That system would most likely have been in full motion trimming down, as it is designed to do, during this go around.
So how are we supposed to discover a runaway trim? After 6 hours flight time, at night, ****ty weather, tired and turbulence. Very easy to miss.
Never heard about a runaway trim on the NG, though.

STS is another weird Boeing system. Most of the time I end up trimming to cancel out this system dueing take off. Why they put it in there, I have no idea.

Iceing? I don't buy that, partly because of that effective HS, and I don't know of a single incident with ice on the HS on a 737. The frozen stab on that Norwegian NG is not applicable for the conditions this FZ aircraft was in. 6 degrees on ground means a lot of any ice would have melted off during approach.

Accidently trimming by hitting the wrong switch?
A. The auto pilot will disconnect if you use stabilizer trim.
B. If you fly manually, you will feel this after a few turns of the trim wheel. It would be very weird if you continue to trim and fight the out of trim with the yoke at the same time.
Unless the PM was trimming? That would be weird too.

There has been a number of near accidents related to level off and pitch down during level off (with aircraft without any defects) as pr the Boeing magazine somebody copied from way back in this thread.

[Util BUS]

Speed Trim System 
The speed trim system (STS) is a speed stability augmentation system designed to improve flight characteristics during operations with a low gross weight, aft center of gravity and high thrust when the autopilot is not engaged. The purpose of the STS is to return the airplane to a trimmed speed by commanding the stabilizer in a direction opposite the speed change. The STS monitors inputs of stabilizer position, thrust lever position, airspeed and vertical speed and then trims the stabilizer using the autopilot stabilizer trim. As the airplane speed increases or decreases from the trimmed speed, the stabilizer is commanded in the direction to return the airplane to the trimmed speed. This increases control column forces to force the airplane to return to the trimmed speed. As the airplane returns to the trimmed speed, the STS commanded stabilizer movement is removed. 
STS operates most frequently during takeoffs, climb and go-arounds.

Conditions 
for speed trim operation are listed below: 

• Airspeed between 100KIAS and Mach 0.5 

• 10 seconds after takeoff 

• 5 seconds following release of 
trim switches 

• Autopilot not engaged

• Sensing of trim requirement

[langleybaston]

QUOTE

I agree, framer. I'm not out, but it's a site for 'professional pilots'. There's some interesting stuff as the awful facts about another accident surface. Would be appreciated if those not flying planes for a living looked and didn't touch.

That is arrogant. The contributions by qualified meteorologists on this thread have been significant. I doubt if many of us fly planes for a living.

[framer]

Right....I'm in :)
Bring on the Met guys and anyone with relevant information.
TWT made a post ( and then deleted it) confidently stating that the trim switch is only on the Captains side. He didn't say " I think the trim switch is only on the Captains side" or " Maybe the trim switch is on the Captains side" , he confidently told all and sundry that this is the case.
Now.....nobody who has ever held a type rating on a 737 would ever say that or think that yet here he is skewing a conversation, holding zero responsibility, mis advising the Met guys , the Airbus guys, the ATCO's , the media, everyone ( except those 737 rated) because for a milli second he gets to feel like he is an expert. For every half decent post there are six posts that are complete rubbish/ ego trips/ lies.
The reason this is frustrating is because a few years ago this site was full of professionals who gave insightful opinions, now, it is not.
The money will dry up when good info can no longer be gained from this site.
PS I wasn't having a go at you Mr Snuggles, just the guy who fed you duff gen.

[aox]

could flight level 80 be 800 meters? That would be around 3200ft-3500ft which is the altitude they reached during both go-arounds...

Unlikely, unless something is affecting mental arithmetic.

And 3281 feet is 1000 metres, not 800

[Green-dot]

No lady yelling STABILIZER MOTION, STABILIZER MOTION, STABILIZER MOTION on the NG.

So how are we supposed to discover a runaway trim? After 6 hours flight time, at night, ****ty weather, tired and turbulence. Very easy to miss.

Unless the PM was trimming? That would be weird too.

The good old DC-8 Trim in motion "wrong answer" buzzer would have had a function in identifying such conditions . . . .

[PT6Driver]

PBY and others.
Oldkingcoal, votex et al have repeatedly said that Fly dubai do NOT do dual autopilot approaches, they would NOT have done a single autopilot approaches. In this case the captain (That's the guy who on this occasion was in the left seat) would have flown the approach manually using the HUD.
Therefore there was no autopilot go around. At the very least the initial part of the go around would therefore have been flown manually.
Please take this into consideration when putting forward your non type rated theories.

From the possible translation theories could this strange word refer to the VS thumbwheal. Which would match the description of the "thing on a computer mouse"?

KS
Could you expand on the statement refering to stall and wing drop by MAK?

[Util BUS]

Strange. Just can’t get my head around it either. Cannot imagine them missing something so serious or not being able to pull out of the dive. Regardless of stabilizer position, the B737 still has manual reversion and enough force should have gotten them out of a dive.

As said above, the trick is to find out why they went from close to 15-20 deg nose up to nose down. All I can think of is that now that the Windshear warning was gone they were trying to get back to a clean configuration.

In context I guess when they departed they had the following:

APS 42.5T
Traffic Load 5.5T
Take-off Fuel: 20.0T
TOW 68.0T
Burn to 1st App 9.5T
First approach 58.5T
Further 2.5h hold 6.5T
2nd Approach 52.0T

Vref 30 at 52T is approx. 132 kts

Limit speeds:

F1: 250
F5: 250
F10: 210
F15: 200
F25: 190
F30: 175
F40: 162

So it looks like they would have still had a 40 margin for flap overspeed during the initial go around. When the windshear warning stopped they would have selected flaps 15 and gear up. Followed by flap retraction in accordance with the bug. Looks like it never got to that.

[ManaAdaSystem]

Manual reversion?
They did not fly without hydraulics as far as I know.

Yes, the trick is to find out why they crashed.

[Otto Throttle]

The crash videos seem to be consistent with a lighting pattern indicating gear down. This may not have been a regular go around, but windshear escape, the altitude from which it started being consistent with system parameters. It could be possible to forget the gear however (been there, done that), especially IF fatigue is a factor, but I wouldn't expect to see that level of performance with gear down at standard GA thrust.

As for Colgan, that was a crash where the captain had repeatedly failed check rides and so therefore has to have been of suspect ability, and a relatively inexperienced FO, both of whom managed to grossly mishandle the situation. That flight crew profile does not appear to match this incident, but that does not mean any human is immune from making mistakes.

[Lost in Saigon]

In reference to some of the points mentioned above.

On the ATC recording at time 5:58 the crew advise in case of a go-around they will climb to FL80.

The B737-800 FCOM 2 Section 4.20.2 states that the auto pilot will disconnect automatically in case of:

Pushing TO/GA on a single channel approach:

- Below 2000 ft RA
- with flaps not up or
- G/S engaged

Also if they had a STAB Trim runaway it would have been obvious to the crew from the sound of the trim wheel and would be quickly discovered by the investigators given the associated sounds on the CVR.

I would be much more inclined to believe it was due to tailplane icing such as in the
October 1994 American Eagle ATR72 crash which was further investigated in NASA's Tailplane Icing program of March 2000.
I agree. Tailplane icing is the most obvious answer to the perplexing question as to why did the nose pitch down so rapidly even though the pilot was pulling back on the control column.

[threemiles]

At 6:03 the controller responds to their intention to climb to FL80 with "Roger" although this is non-standard terminology I would have accepted it to me "Approved" in that part of the world.


Roger is very much so ICAO standard terminology and means, I acknowledge receipt of your transmission.

It is NEVER a clearance, in no part of the world.

The required clearance would read: In case of go around climb to and maintain FL80.

I hope your airborne appearance is limited to simulator skies, otherwise I suggest to revisit an ICAO phraseology RT lesson asap.

[ITman]

Spotted this on Channel News Asia..

"Pull up!": Russian TV airs final words of pilots in Flydubai jet crash - Channel NewsAsia (http://www.channelnewsasia.com/news/world/pull-up-russian-tv-airs/2638914.html)

[Sciolistes]

Threemiles,
Data are from Air Data Computer. No data are calculated outside the airframe. Most likely recorded FDR data will appear 100% the same, albeit at a higher sampling rate.
MSL altitiude is from GPS (height over ellipsoid) and transmitted as an offset to QNE baro altitude.
Open remains the source of VS. There are two options: GPS or Baro. Depends on the supplier of the avionics.
Thanks for that. Great explanation. BTW, have you noticed how on the TCAS display how the an aircraft ahead can jump around in azimuth and also how a level aircraft ahead can show alternating climbs and descents with changes of around 100ft?

[Aluminium shuffler]

73 only used baro and radalt inputs, nor GPS altimetry.

[LlamaFarmer]

At 6:03 the controller responds to their intention to climb to FL80 with "Roger" although this is non-standard terminology I would have accepted it to me "Approved" in that part of the world.
Roger is very much so ICAO standard terminology and means, I acknowledge receipt of your transmission.

It is NEVER a clearance, in no part of the world.

The required clearance would read: In case of go around climb to and maintain FL80.

I hope your airborne appearance is limited to simulator skies, otherwise I suggest to revisit an ICAO phraseology RT lesson asap.


I accept that, but therefore they acknowledged the intention to climb to FL80.

The fact they didn't deny a clearance to FL80 could have been interpreted as a clearance-by-omission.


I've never flown to Russia, but I know plenty of people who have, and their non-standardness extends beyond just units. Phraseology is not always from the ICAO textbook, and in this instance, whilst it should not be accepted as a clearance, it is very possible that the controller meant it as an approval.

[Kulverstukas]

@pt6driver

"At an altitude of about 900 meters and a speed of more than 400 kilometers per hour, the plane banked left and hit the ground in the left edge of the runway almost vertically, at an angle of 60 degrees," Tass news agency quoted a source with Russia's air transport authorities as saying.

From MAK sources after reading FDR around March 20.

[Ian W]

o how does the plane go from 0° to 45°-90° bank angle without lateral deviation ?
Forward motion provided, a passenger jet makes a turn when it banks, isn't it ?

OK it could have stalled, but at the speeds reported it appears unlikely. However, if there was a lot of nose down trim any lateral deviation due to the bank angle will be counteracted by the nose down pitch caused by the full nose down trim

[Rananim]

Accidently trimming by hitting the wrong switch?
A. The auto pilot will disconnect if you use stabilizer trim.
B. If you fly manually, you will feel this after a few turns of the trim wheel. It would be very weird if you continue to trim and fight the out of trim with the yoke at the same time.
Unless the PM was trimming? That would be weird too.

The little CVR data(if true) we have implies one pilot did something the other didnt exactly like.The PM makes a call to ATC(about diversion so probably more than momentary call) but he mistakes the transmit switch for the trim switch.If this mistake was made,he would be trimming down when he thought he was transmitting to ATC.Theyre opposite each other on the stick.
If the PF is flying manually there is no AP disconnect wailer to let the PM know the error of his ways.Environmental noises(wshear rain) may camouflage the trim wheel noise.PM isnt holding the stick so is unaware of the forces on the stick that he is unintentionally creating.Only PF.If this situation is not corrected almost instantly by PF( trim cutout switches/remove PM's hand from stick) then at this altitude,it could be fatal.
Icing is not the cause.99% sure of this.
All conjecture but this had to be catastrophic flt control problem.Whether pilot induced or not.Vmca encounter also possible.

[PBY]

PT6driver quote:PBY and others.
Oldkingcoal, votex et al have repeatedly said that Fly dubai do NOT do dual autopilot approaches, they would NOT have done a single autopilot approaches. In this case the captain (That's t

PT6Driver, it has already been said on Russian TV that they DID USE the autopilot.

[Lost in Saigon]

Icing is not the cause.99% sure of this.

Why are you 99% sure the cause is not icing?

[aterpster]

Green-dot:

The good old DC-8 Trim in motion "wrong answer" buzzer would have had a function in identifying such conditions . . . .

No one wore noise-cancelling headsets during my career. Now, most crews do. I have to wonder in some circumstances if that could create a possible safety problem because of greatly diminished ability to hear flight deck aural advisories or warnings? Or, other pertinent sounds?

[Old Boeing Driver]

Good question about the noise-canceling headsets.

I used some late in my career and, as I recall, one could still hear pertinent warnings and sounds. This was not on the 737.

The trim wheel and sound on the 737 was pretty hard to miss. Visually as well.

Are there any 737 guys using those types of headsets who could comment on this?

[Capt. Inop]

The HS on the NG is a very powerful device. If it is grossly out of trim, the elevators are not big enought to get you out of the situation.

Indeed.
And on approach, and goaround where you hear the trim move around alot wold you catch a runaway trim situation in time to recover?

[PBY]

Why do you guys talk about runaway trim situation, when it is a predictable situation, that the autopilot will trim hard nose down, if it is trying to level out against the "will" of full power set? And than disconnects due to the pitch down max attitude exceeded?
It is a situation that can be clearly predicted. If you are not caught by surprise.
Somebody, who has an access to a B737 sim should try the "high energy goaround in the sim"

[helen-damnation]

In relation to the noise cancelling headsets; how do we know if they were using them to cover both ears?
Many people, self included, have one ear only partially or totally uncovered. It's usually the side closest to the centre console, ie right ear for Capt, left for FO.

[Avenger]

The "Russian Translation" if accurate implies the stabiliser was running away, either uncommanded or by human intervention Aka all the theories voiced herein. Something about a " fin switch" can only relate to the stab trim cut outs by the throttle quadrant, part of the memory items, or the black guarded cut out switch at the rear of the pedestal used by engineers ( clutch). I agree if I was flying and the other person was trimming down I would quickly notice this and firmly holding the column would stop this, so thats more than confusing to say the least. Runaway stabiliser is trained in the sim and unless the crew were operating with opposed opinions of the events it should be manageable. Its conceivable that with fatigue and hand over of controls for weather conditions a complete lack of SA occurred, sadly.

[PBY]

Interesting incident:

http://www.skybrary.aero/index.php/B752,_vicinity_Gardermoen_Oslo_Norway,_2002_(LOC_HF)

Looks it could be very related

[Capt. Inop]

Well it could, but i don't think so.

(we were nr2 for takeoff after the Icelandair 757 had landed)

[Alice025]

Since the trim switch is a rocker switch, rather than a true switch, is is possible this refers to the trimmer.


Didn't see a comment in Russian language blogs where a Boeing pilot would immediately recognise "knuppel", like "Oh, I know what it is." Russian Boeing pilots seem to have learned their aircraft in English manuals and documentation that don't mention the name. Older generation technicians, though, recalled knuppel.
Can be knuppel was fed to the media because it allows for different interpretations. Or it was obvious for the leak source that it is a perfect knuppel.
I give up.

[txl]

My theory, being a journalist: The "knuppel" might be referring to the German "Steuerknüppel" or just "Knüppel", meaning stick or yoke in an aviation context – as has been mentioned here before. "Knüppel" usually means "club".

The "knuypel" might have made it into the Russian news story by means of miscommunication between a journalist and his/her source. The source here probably being not a Russian, maybe German, and the journalist not well versed in aviation terminology. I write about tech and see something like that happen a lot when it comes to specific technological language.

I'd think the source may have meant something like trim was activated by a switch on the yoke or "ein Schalter am Knüppel" in German.

[level_change]

If this mistake was made,he would be trimming down when he thought he was transmitting to ATC.

You cannot trim down in the NG while there is control wheel backpressure in manual flight. Which is what you would naturally do to keep your attitude.

[Otto Throttle]

I wear a modern ANR headset, with very good passive noise reduction too, and the trim wheel is absolutely clearly audible over all the other flight deck noises. Unlikely that it could be missed (but obviously not impossible) but more likely it simply doesn't attract attention with it being a familiar noise at this phase of flight. A crew would have to recognise the trim running for an unusual length of time, but with other more "urgent" distractions such as weather, or a degredation in capacity due to fatigue, I could accept it as a possibility.

[threemiles]

Lost in Translation?
My theory, being a journalist: The "knuppel" might be referring to the German "Steuerknüppel" or just "Knüppel", meaning stick or yoke in an aviation context – as has been mentioned here before. "Knüppel" usually means "club".

The "knuypel" might have made it into the Russian news story by means of miscommunication between a journalist and his/her source. The source here probably being not a Russian, maybe German, and the journalist not well versed in aviation terminology. I write about tech and see something like that happen a lot when it comes to specific technological language.

I'd think the source may have meant something like trim was activated by a switch on the yoke or "ein Schalter am Knüppel" in German.

Also the "wind scissors" are a "wind shear", because in German "schere" means "scissors" and "windscherung" means "wind shear".

The tv report says the commander disconnected the AP at the time of a wind shear and he had used the knüppel (rocker switch) before to stab trim down or stab trim happened for unknown technical reasons.

[Kulverstukas]

The "knuypel" might have made it into the Russian news story by means of miscommunication between a journalist and his/her source.

It definitely is, but not because source was german, but probably because source was from military.

[fireflybob]

I wear a modern ANR headset, with very good passive noise reduction too, and the trim wheel is absolutely clearly audible over all the other flight deck noises. Unlikely that it could be missed (but obviously not impossible) but more likely it simply doesn't attract attention with it being a familiar noise at this phase of flight. A crew would have to recognise the trim running for an unusual length of time, but with other more "urgent" distractions such as weather, or a degredation in capacity due to fatigue, I could accept it as a possibility.

Perhaps we should also keep in mind that when humans become maxed out with a task the first sensual input the brain deletes is that of hearing.

[LLuCCiFeR]

More from RT: https://www.rt.com/news/337270-emirates-pilot-brutal-roster/

[Alice025]

As to Roger, I think it meant "Got it" since it was said by the tower.


(Non-aviation - related Russian might think "Roger" is a flashy foreign way to say "Bye, end of talk." Influenced by video movies watched in 1990-s with police stories and various American police hoovering helicopters. They always said that in the end of radio talks.)

[Avenger]

If one recalls the NAS incident in Kittila 26/12/2012,
Ok this was due to icing fluid accumulation , however, given the conditions it may be worth a thought..although in this event the crew were working in harmony where as it would "appear" in the Rostov event there was some confusion.

[silvertate]

But in B737 or A320 which has THS, you do not necessarily trim out the force. As a matter of fact, you can have the control column in neutral position with no force, while the THS is being trimmed hard forward and not feel anything on control column.

Why do people who hace never flown the 737 write drivel, and muddy the waters? They really should go back home, I can hear Muma calling.

[aox]

Also the "wind scissors" are a "wind shear", because in German "schere" means "scissors" and "windscherung" means "wind shear".

I think that was quite easy to realise, thanks. Also in English there are ranges of similar implements called shears, for garden use, sheep, surgery, dressmaking, metalwork ...

It appears that some of these translations have been done by automatic means, perhaps sometimes two or three times successively.

[flyingchanges]

"B737 trim does not work the same as small aircraft trim. In small aircraft you trim out the forces. But in B737 which has THS, you do not necessarily trim out the force. As a matter of fact, you can have the control column in neutral position with no force, while the THS is being trimmed hard forward and not feel anything on control column."





Actually to some degree, this is true. You still trim out the control column forces, but for the most part, the yoke remains centered when trimmed. In a light ac, you are changing the position of the elevator and yoke with trim changes.

[Old King Coal]

Confirming what ManaAdaSystem has said, and for the avoidance of all doubt, flydubai policy is that dual-autopilot approaches are not approved and must not be utilised. Furthermore, the use of dual-autopilots is part of the 'data monitoring' process and it is therein possible (retrospectively) to determine if dual-autopilots have been utilised. The last FZ pilot who took it upon their-self to utilise dual-autopilot's for a landing was sacked for wilful breach of company SOP. Whilst a single autopilot might be utilised for part of an approach, at some point during said approach, that autopilot (and the auto-throttle) is required (SOP) to be disengaged and accordingly all landings (regardless of them being either category I/II/III) and go-arounds (in flydubai) are therein flown manually; albeit always possible for a pilot to re-engage an autopilot and the auto-throttle at some point during a go-around (subject to all the normal engagement criteria).

[AtomKraft]

Let's say they went around for Windshear. As they pull up, they meet a greatly increasing head wind component.

The GA starts, and from what I've read, a 737 pilot has to push, and trim forward, especially with a light a/c.

So: pushing forward, would make the THS trim nose down on a 737? (it does on the non-Boeing type I fly).

Full power causes nose to rise due to engine thrust line below the wing, so more nose down trim to stop it rising too much.

Extra energy from increasing HWC, would tend to need nose down trim as ROC increases as aircraft meets rapidly increasing headwind.

Generally high ROC with full power and while hand flying might cause PF to trim forward to reduce the ROC.

Reaches GA standard level off alt.

As they level off, thrust comes right back, reducing that upward thrust moment, and causing the a/c to go nose down with THS already almost full forward, nose down.

They are IMC, so miss the gravity of the situation for a few seconds (some distraction or other).

Big ROD develops.

Come out the bottom, see the ground,but too late.

[Mr Optimistic]

SLF: where does the bank angle come from and would that be hard to notice?

[anson harris]

Great article here - explains what that pesky stab trim actually does. When I saw the first video of the high ROD crash following the beginnings of a GA, my mind turned immediately to this, as well as the Bournemouth ThomsonFly incident:

Roger-Wilco | Do you really understand how your trim works? (http://www.roger-wilco.net/do-you-really-understand-how-your-trim-works/)

[Chesty Morgan]

if your nose is stuck way down(stab down)last thing is to add thrust which will just add the downward energy,speed increase and even harder to pull out.
Adding thrust for pitch up works wonders when in trim condition not in that scenario.
You need to get back to the speed at which it was in trim,reduce thrust which will allow the elevators more control over the jammed stab.(less manual force required).
I've not read the rest of the thread so forgive me if this has been answered already but you're not exactly right.

As a 737 air tester it is standard to brief the use of power to recover from an out of trim nose low attitude. Especially in a low speed scenario it is one of the first things I would consider (and it works) if my primary pitch control isn't as effective as it should be.

Reducing thrust with an aircraft out of trim nose down is the last thing you want to do.

[silvertate]

. B737 trim does not work the same as small aircraft trim. In small aircraft you trim out the forces. But in B737 which has THS, you do not necessarily trim out the force. As a matter of fact, you can have the control column in neutral position with no force, while the THS is being trimmed hard forward and not feel anything on control column.

Actually to some degree, this is true. You still trim out the control column forces, but for the most part, the yoke remains centered when trimmed. In a light ac, you are changing the position of the elevator and yoke with trim changes.

Firstly, not all light aircraft are the same. The trimmer on a Piper Cub is exactly the same as the 737. And yes, the stick or cc will remain in the same place as you trim forwards, but unless you give a mighty pull on the stick or cc, the aircraft is going south very quickly. So you cannot trim without experiencing stick or cc forces.

And regards this incident, I cannot see anyone trimming forward without noticing the loud racket from the antique trim system, or the sudden pitch down, or the large cc forces required to resist this. The only thing that might make sense is a trim runnaway, that was not caught in time. Remember that if the flaps were still extended, the trim runs at double the rate. Not sure why it would not be stopped, as it is easy to hear and see, opposite cc pressure should stop it, and you can grab it with your hand if you don't mind losing a finger.

The mystery remains. The only thing we can perhaps dismiss, is a stall or a Bournemouth thrust pitch, as the VS, altitude and GS plots don't support these scenarios.

[anson harris]

The only thing we can perhaps dismiss, is a stall or a Bournemouth thrust pitch, as the VS, altitude and GS plots don't support these scenarios.

Just to clarify, I'm not saying it was a repeat of Bournemouth, just in the sense that my first thoughts were that it may be pitch trim related.

[dougydog]

FWIW- IMO- This has all the hallmarks of the Bournemouth incident and will be exactly where the authorites would be looking for similarities.

https://www.gov.uk/aaib-reports/aar-3-2009-boeing-737-3q8-g-thof-23-september-2007

[Otto Throttle]

I couldn't see if it had been mentioned yet or not, so perhaps someone can link to it, or our weather guys could shed some light on the wind changes within the altitude band we're looking at. The wind speed has been mentioned at 600m as being somewhere in the order of 60-70 kts if I recall, but not direction.

On the surface, the wind appears to be fairly steady directionally from 250 degrees, so a crosswind from the right (runway 22), so assuming it veers with altitude, it becomes a stronger, more direct crosswind. This doesn't fit with a rapidly increasing headwind scenario, as the change in direction would to a large extent cancel out substantial changes to headwind component. If the direction remains steady however, then that theory holds water.

[Aluminium shuffler]

The BOH and AMS events occurred where elevator authority was exceeded by a combination of trim and thrust couple largely because they ran out of airspeed. The IAS does not appear to have reduced significantly here, so they would have had enough elevator authority unless there was some problem with the elevator.

[Old Boeing Driver]

You may be on to one of the issues. Here is a quote from the Bournemouth report.

"The trimmed position of the stabiliser, combined with the selection of maximum thrust, overwhelmed the available elevator authority."

If the FZ guys got trimmed one way or the other without realizing it (and I find that hard to understand), this could be a similar situation.

[FullWings]

How could the plane have gone from 0° bank angle to 45°-90° bank angle WITHOUT lateral course deviation ?
If the wings were pretty much unloaded, quite easily. You need an angle of attack on the mainplane to generate lift in order to accelerate away from the course you’re on. If the aeroplane was in a bunt, which seems to be the case, it’ll rotate about the longitudinal axis and that’s about it, a bit like an aileron roll. This was not a “balanced turn”.

In some ways this is good evidence that something was seriously wrong in terms of stabiliser and/or elevator position.

[Aluminium shuffler]

Old Boeing Driver, my post above was to counter the previous comments about similarity with BOH. If the speed and altitude vs time graphs are true, then IAS never reached a point where this would have occurred.

For all those speculating about active noise reduction headsets, please leave PPRUNE; you are clearly not pilots and have no idea how ANR works.

[604driver]

Do all FD aircraft have a HUD? Is it SOP to always them?

Personally, I'm not a fan of them at all. Ive seen 2 good drivers, in the sim, stuff it all up during a GA (resulted in the loss of the aircraft), trying to recover from an abnormality, when using the HUD as opposed to doing it the way they had been trained for, for the previous 25 years, where the result was never in doubt and the "event" was a non event.

Whether it will be a factor in this accident, i don't know, but so far on this thread, no one has discussed their limitations and it might just be necessary to start, if indeed a HUD was fitted and used.

[Old Boeing Driver]

I understand. Thanks for your clarification.

Regards.

[Chesty Morgan]

If the wings were pretty much unloaded,

And if they were then a stall is unlikely.

[xollob]

I initially was thinking about the Bournemouth incident also but as someone else said there does not seem to be a low speed event (from the data that has been posted).

If the leaked CVR transcript is accurate and there are shouts of Pull we don't know if it's because of a stall situation or the aircraft being in an nose down situation and still "flying".

Someone said it appears to have got to (perhaps) the MCP altitiute okay, then suddenly it all changes, this could be that it got there okay as one would expect, or it got there converting forward energy converted into vertical climb rate and then stalled (data so far doesn't seem to suggest that).

So left wondering what would cause rapid descent

Stall
Uncommanded nose down
Fight control problem

I wonder if during the go around if the a/p was engaged and because of the crazy rate of climb the A/P shoved in loads of fwd trim to try and reduce the climb rate, or the pilots did.

Then once the thrust came off reaching altitude they are now left without the pitch couple to the thrust and aircraft nose naturally comes down, but the stab is now fully forward also, with very little Sky available to try and resolve it ?

[oicur12.again]

Never flown a Boeing but many years flying various FBW bus's.

Reading the comments here about pilots mishandeling trim and speed and pitch difficulties during GA makes me think more airlines should be buying A320's instead of 737's.

[parkfell]

O12.again

Except perhaps when ALTERNATE LAW comes into play?

[45989]

Fly both types all the time.One needs to know the differences. Preference, Boeing. your arse tells you whats going on. Airbus is by numbers only

[fireflybob]

Having followed this thread and read all the posts my feeling is that this is more than a mishandled go around, whether or not the autopilot was in use.

Even fatigued pilots know how to fly a go around and these pilots were not beginners.

[Old Boeing Driver]

I originally thought this might be a high speed event due to the light airplane with loads of power.

However, I think Dave Reid pointed out that one of the last GS plots was 185KTS. As they started down, GS increased.

With very little info, we are all still guessing and speculating, but the Bournemouth incident made me think of another scenario.

If it was me, I would have gotten the automation back on after the GA and clean up due to the fatigue factor and I would want to monitor all that was happening.

It looks like they had a high ROC right after the GA. If the PF had turned on the automation, and idled the power, he may have be looking at other things well into a speed degradation (like Bournemouth) until the plane was actually stalling.

In the Bournemouth incident they had a 44 degree pitch angle. The same or more could have happened here, and maybe when the break occured, then they may have had the power all in and never touched it as they tried to recover.

IMHO, it would be difficult to get descending at 18,000 FPM in just a few seconds without the power helping.

Just another scenario to consider. We'll know when the FDR and the full CVR info is published.

[737er]

If they engaged the AP on the miss it's the AP trim motor that runs the stab. It runs slowly and it is normal to hear it during that circumstance. It would be trimming nose down almost certainly. If they engaged it during a highly dynamic situation the system can't keep up often times and begins pitch oscillations. If they turned off the AP in response to a pitch up oscillation with the AP still trimming nose down, they might have got a big pitch over surprise as the AP disengaged. We all need more information obviously. Just throwing this out there.

[mommaklee]

I couldn't see if it had been mentioned yet or not, so perhaps someone can link to it, or our weather guys could shed some light on the wind changes within the altitude band we're looking at. The wind speed has been mentioned at 600m as being somewhere in the order of 60-70 kts if I recall, but not direction.

On the surface, the wind appears to be fairly steady directionally from 250 degrees, so a crosswind from the right (runway 22), so assuming it veers with altitude, it becomes a stronger, more direct crosswind. This doesn't fit with a rapidly increasing headwind scenario, as the change in direction would to a large extent cancel out substantial changes to headwind component. If the direction remains steady however, then that theory holds water.The best wind data available is from the Airport's Sounding taken at ~0000 UTC 19MAr16. Here are the data for the relevant lower altitudes:

http://i.imgur.com/b3gNTej.png

Speed drops near the surface in what one would expect would in power-law fashion, with increasing gustiness, while backing 'gradually' from ~westerly to ~southwesterly.

I took a stab at calculating head winds and cross winds for the second GA:

http://i.imgur.com/O4NekMc.png

with an corresponding GPS altitude chart which also includes a stab at IAS:

http://i.imgur.com/gcKtoko.png

I hope that helps some.

(Sounding data: University of Wyoming - Radiosonde Data (http://weather.uwyo.edu/cgi-bin/sounding?region=mideast&TYPE=TEXT%3ALIST&YEAR=2016&MONTH=03&FROM=1900&TO=1900&STNM=urrr))
http://imgur.com/b3gNTej

[Old Boeing Driver]

Good points. It's been a long time since I flew the 737, but can remember a few surprises as you mentioned.

I really want to know how it happened so fast.

Regards.

[mommaklee]

While I am posting images, I fixed up a comparison of the two GA attempts, with time zero being the first indication of pitch up in the single station ADS-B data from FR.

http://i.imgur.com/guCYzNK.png

and

http://i.imgur.com/Yc42PGW.png

FWIW.

The difference between 17 and 30 sec is very curious to me, but I don't know enough to speculate as to what it indicates.

[Old Boeing Driver]

How did you calculate the IAS?

[Sunamer]

how does the plane go from 0° to 45°-90° bank angle without lateral deviation ?
It would be possible, when it is unloaded...
In other words, when AoA is such that it gives minimal lift. In that situation, there will be almost no lift vector and, hence, nothing to pull the AC into the turn.

[mommaklee]

How did you calculate the IAS?

I converted vertical speed from FPM to knots, then calculated the hypotenuse (vector if you will) speed. A year ago or so I ran across a reference which gave as an estimation between TAS and IAS: (TAS/(height_in_ft/50000+1)) [don't have the link handy as it is on my work computer]. From that I roughly estimated at the altitudes in question TAS~= the above calculated speed + head wind. All quite rough, not having an E6B.

If I can improve upon that given the FR24 and sounding data, I would love to do so! :)

[Yankee Whisky]

"B737 trim does not work the same as small aircraft trim. In small aircraft you trim out the forces. But in B737 which has THS, you do not necessarily trim out the force. As a matter of fact, you can have the control column in neutral position with no force, while the THS is being trimmed hard forward and not feel anything on control column."





Actually to some degree, this is true. You still trim out the control column forces, but for the most part, the yoke remains centered when trimmed. In a light ac, you are changing the position of the elevator and yoke with trim changes.






Trimming the stabiliser versus the elevator is the difference.

[Old Boeing Driver]

Well done and explained.

I think your calculations are pretty close.

They accelerated from 200KTS to 300KTS in about 7 seconds.

Airspeed, Altitude, and ideas are the basic requirements of flight. You have to have 2 of the 3 to be relatively successful.

It looks like they ran out of ideas and altitude too late.

Thanks for your posts.

[RatherBeFlying]

Mommaklee,

Once you're in a steep climb or descent, there's an extra tangential "headwind" component which has to be brought into the headwind component calculation.

IAS and TAS will be along the hypotenuse while GS is generally given along the horizontal component.

[Ranger One]

http://i.imgur.com/gcKtoko.png



What might be enlightening is to annotate that plot with the reported (and time-correlated) CVR remarks...

R1

[CONSO]

Just what we do NOT need is an article from the NYT- via seattle times

Transcript indicates cockpit decision may have doomed FlyDubai jet that crashed in Russia | The Seattle Times (http://www.seattletimes.com/nation-world/transcript-indicates-cockpit-decision-may-have-doomed-flydubai-jet-that-crashed-in-russia/)

By IVAN NECHEPURENKO (http://www.seattletimes.com/author/ivan-nechepurenko/)
The New York Times

MOSCOW — An error by a crew member committed during adverse weather may have been responsible for the crash of a passenger jet last week in the southern Russian city of Rostov-on-Don that killed 62 people, a report broadcast by Russian state television said.
The Rossiya-1 television channel said it had obtained a transcript of the pilot interactions a minute before the FlyDubai Boeing 737-800 passenger jet nose-dived to the ground, killing all the passengers and crew members. A source in the investigative commission with access to flight recorders provided the TV channel with the transcript.



Goes on to claim wrong button was pushed re trim of HS or some such- eg pilot error..

[Rananim]

You cannot trim down in the NG while there is control wheel backpressure in manual flight. Which is what you would naturally do to keep your attitude.

Yes.But activating the under-floor cutout feature requires a pronounced opposing stick force.
If accidental nose down trim went unchecked for longer than say,several seconds at that altitude,they wouldnt have had a lot of room to play with.It remains a possible scenario.More data is needed to narrow it down.

[Old Boeing Driver]

This is a quote from the NYT article.

"The television channel cited experts who suggested that by turning off the autopilot, the pilots were trying to pull the plane back to a horizontal position. But at that moment, a stabilizing fin at the jet’s tail was switched on.
With the fin activated, “the elevator is no longer working and the plane practically does not react to the pilot’s control panel,” the report said. The channel suggested that the pilot could have accidentally hit the button that activated the fin because of his reported “chronic fatigue.”"

Can anyone translate this English into English?

What could they be talking about..."a stabilizing fin at the jet’s tail"?

Edit: maybe the HS?

[eduelp]

Never flown a Boeing but many years flying various FBW bus's.

Reading the comments here about pilots mishandeling trim and speed and pitch difficulties during GA makes me think more airlines should be buying A320's instead of 737's.

Yes, because the best solution for pilots loosing basic stick and rudder skills is to just put a computer to do that for them... And we know how that turned out with AF447...

[CONSO]

What could they be talking about..."a stabilizing fin at the jet’s tail"?

Edit: maybe the HS?

Just a SWAG from this SLF. Much confusion in many- MOST - press accounts between what is the ELEVATOR and what is the HORIZONTAL STABILIZER.

[wongsuzie]

The NYT article doesn't make any sense.

They can't diffenciate between 'fin',HS and stab.

[mommaklee]

What might be enlightening is to annotate that plot with the reported (and time-correlated) CVR remarks... I am not particularly confident in the veracity and completeness of the CVR transcription as provided in the media thus far, It is what we have for now and worth discussing. With that...

http://i.imgur.com/8EJA0AT.png - IAS anot

(I updated the IAS calc using the method suggested here: http://www.pprune.org/tech-log/449863-tas-ias-pa-oat.html#post6419000 -- Found that just after my reply to OBD. (Thank you by the way for your kind words. They were very much appriciated :) ))

Also, for what it's worth, remember the cloud base was in the vicinity of 1700 ft above *MSL* (I am using GPS altitude in these plots).



These annotations may also be helpful here:

http://i.imgur.com/9fwOEdQ.png

with

http://i.imgur.com/guCYzNK.png

for comparison.

The period ~18 to ~30 seconds after the initiation of the second GA in the last two has stood out to me over the past couple of days as interesting, but I don't know what to make of it myself.

[klintE]

I think that flight path of the a/c with HS in full down position (as claimed on RU tv) should be different. Definitely not a straight line - like we can observe in video showing last few seconds before impact.
But this scenario maybe coul'd fit when to presume that pilot counteracted with full column pull back.

On the other hand, to move stabilizer to full down position (f.e. with elec. stab trim) you need time - it's 30 maybe 40 wheel turns from nose up position. And on the presented ADS-B data we can see rather fast pitch drop (from +10 to -30 in 7 sec.)

So something still is very unclear in this case

[CodyBlade]

Definitely not a straight line - like we can observe in video showing last few seconds before impact.
But this scenario maybe coul'd fit when to presume that pilot counteracted with full column pull back.

Exactly,yoke at full aft position at that speed and nose doesn't even raise 1 inch?...

[Otto Throttle]

The press and some contributors seem to be making the assumption that the "Don't do that" on the CVR corresponds to some catastrophic pilot action. I'm struggling to think what could have been changed or deselected at this point in flight which was instant and irreversible.

I'm of the opinion the aircraft was still in a landing configuration at impact, suggesting a wind shear event (consistent with starting altitude and subsequent performance). This means flaps would have initially been set at flap 30. Even if the alleged pilot action was flap retraction, the initial motion would have been an improvement in performance, not a sudden stall. Any further retraction may be an issue, but it is not instantaneous, and given the apparent speeds, very unlikely at this weight to have left the aircraft at a dangerously low speed. A flap mis-selection could also very quickly be reset. Personally, I don't buy it.

[Lost in Saigon]

I think that flight path of the a/c with HS in full down position (as claimed on RU tv) should be different. Definitely not a straight line - like we can observe in video showing last few seconds before impact.
But this scenario maybe coul'd fit when to presume that pilot counteracted with full column pull back.

On the other hand, to move stabilizer to full down position (f.e. with elec. stab trim) you need time - it's 30 maybe 40 wheel turns from nose up position. And on the presented ADS-B data we can see rather fast pitch drop (from +10 to -30 in 7 sec.)

So something still is very unclear in this case


Exactly,yoke at full aft position at that speed and nose doesn't even raise 1 inch?...
I believe the full aft yoke was totally ineffective because the tailplane had stalled due to ice accumulation.

My reasoning is due to two simple facts:

1) The aircraft flew in icing conditions for an unusually long period of time.

2) The aircraft behaved as if the tail was stalled.

Sometimes the simplest answer is the correct one.

[porterhouse]

Sometimes the simplest answer is the correct one.
There was never an accident with tail icing on the 737 so definitely it is not the simplest answer. Yes, simplest answers are usually closest to the truth, the simplest answer is totally mishandled go-around, it happened numerous times and it actually happened in Kazan, Russia barely 2 years ago. And all the symptoms were practically identical. Tail icing doesn't make sense for one simple reason - it would be at its worst during the approach with all flaps fully extended, not the go-around.

[Alientali]

"The television channel cited experts who suggested that by turning off the autopilot, the pilots were trying to pull the plane back to a horizontal position. But at that moment, a stabilizing fin at the jet’s tail was switched on.
With the fin activated, “the elevator is no longer working and the plane practically does not react to the pilot’s control panel,” the report said. The channel suggested that the pilot could have accidentally hit the button that activated the fin because of his reported “chronic fatigue.”"

Can anyone translate this English into English?

What could they be talking about..."a stabilizing fin at the jet’s tail"?

Edit: maybe the HS?

Never mind the b...s. These are the same people who translated the alleged CVR materials into Russian, had actors read them out against the background of artificially added "airplane-y" noises, and put the whole thing on air.

Here's my translation of what was said... for what it's worth (=nothing).

"[After the decision to go around], the aircraft was gaining altitude for 40 seconds. Not yet having reached the planned FL, the captain disconnected the autopilot for reasons unknown - possibly because of wind shear (which they call "atmospheric scissors")/rough air. It is then that the plane began to dive.

After the autopilot disconnect the captain tried to coax the plane into horizontal flight, but after the dive has begun, THE VERTICAL STABILIZER SWITCHED ON. As a result, "руль" [the word they use is actually more like "steering wheel", not "elevator" - which is what they must have meant?] stopped functioning, and the crew didn't have enough time to figure out what happened.

Why the stabilizer turned on is unclear. It is activated with a button which the pilots call "knuppel". In rough air, the captain, while switching to manual flight, might have accidentally touched it without noticing, due to chronic fatigue. Otherwise, it's some unheard of surprise on part of flight automation systems."

:oh:

[Lost in Saigon]

There was never an accident with tail icing on the 737 so definitely it is not the simplest answer. Yes, simplest answers are usually closest to the truth, the simplest answer is totally mishandled go-around, it happened numerous times and it actually happened in Kazan, Russia barely 2 years ago. And all the symptoms were practically identical. Tail icing doesn't make sense for one simple reason - it would be at its worst during the approach with all flaps fully extended, not the go-around.

You make a good point, but all the other theories make even less sense as we know the pilot said "Pull, Pull, Pull" and yet there was no sign that all that pulling did anything to stop the aircraft from violently nosing over.

Tailplane icing is very rare. I doubt we know everything there is to know about it.

If the tail did stall during the late stages of the go-around, this accident could be a game changer.

[olasek]

the pilot said "Pull, Pull, Pull" and yet there was no sign that all that pulling did anything
Have you by any chance heard of laws of Physics?
Do you know what aircraft inertia is?
They had about a few seconds, regardless of the amount of pulling in the remaining 3-6 seconds there would be absolutely no practical difference in aircraft trajectory.

If the tail did stall during the late stages of the go-around, this accident could be a game changer.
If, if, if, could ....
There are lots of ifs and coulds, not the simplest answer you are looking for.

[klintE]

Actually, THS is moving pretty fast when actionated from trim switches, see video posted earlier
https://vimeo.com/34501723

Sorry but this vid seems not to be real, it is just physically impossible IMO
I mean that abrupt changes in angle (units), c'mon
https://www.youtube.com/watch?v=rxPa9A-k2xY thats how it works in fact


I doubt PF accidentally trimmed at full ND. That's impossible; just imagine pilot experiencing impoderability, loose objects on the ceilling and he continuing with trim down.


Yes and I consider a specific dual input - trim from the right seat and yoke pull back from the left.(Or vice versa) And right seat possibly in the state of spatial-D
I know that maybe sounds bit crazy but looking at how that flight ends tells me that something abnormal happened there.

[Chesty Morgan]

Klint, you are correct regarding the first video.

The correct ratio is just under 10 turns of the trim wheel per unit of trim.

[737er]

Main electric trim has two speeds and that video is at the slow speed. (Flaps retracted)

The main electric trim is about twice as fast as the video shows with the flaps in any position but up.

The video does approximate the speed at which the autopilot trim motor runs with flaps up or down.

[framer]

The video does approximate the speed at which the autopilot trim motor runs with flaps up or down.
Not quite.The autopilot trim system also runs at two speeds ( high or low) depending on the flap position.

[Ranger One]

The press and some contributors seem to be making the assumption that the "Don't do that" on the CVR corresponds to some catastrophic pilot action. I'm struggling to think what could have been changed or deselected at this point in flight which was instant and irreversible.

So am I.

I would just caution against reading too much into those words; I'm surprised that no-one has pointed out one obvious possibility; that he was talking to the aircraft - hey it happens; some of us do anthropomorphize a little... it will depend on the intonation and correlation with FDR events.

R1

[Aluminium shuffler]

That would be my assumption, that he was talking to aircraft, not the other pilot. Without seeing control positions and forces, though, it is anyone's guess. A stripped or sheared trim drive like the Air Alaska accident (DC9 or derivative, I think) seems plausible.

[de facto]

Chesty,
a 737 air tester it is standard to brief the use of power to recover from an out of trim nose low attitude. Especially in a low speed scenario it is one of the first things I would consider (and it works) if my primary pitch control isn't as effective as it should be.

Especially in a low speed scenario are the key words id say.
If the combined forces of stuck down stab and high speed (due to GA thrust),i would think that releaving the forces by reducing thrust would help.
It seems that,getting back to intrim speed as early as possible,using the elevator is quite effective.
If both stab and elevator are jammed then thrust is the only help,,increase thrust nose goes up,reduce thrust nose down...
Wouldnt you say?

[FGD135]

Hi all,

My first post on this subject. Firstly, thanks to mommaklee for his outstanding work with the data.

Sorry but this vid seems not to be real, it is just physically impossible IMO You would have to be correct, klintE. The accompanying text to that video says this: Genuine Boeing 737-300 series throttle quadrant converted for flight simulator use. That would be "flight simulator" as in the kind that enthusiasts make at home, that are based on Microsoft FSX!

This accident is strikingly similar to the Tatarstan Airlines Flight 363 B735 that crashed at Kazan, Russia in 2013. This is the graphic of the vertical profile in that accident:

http://upload.wikimedia.org/wikipedia/commons/thumb/b/b5/Tatarstan-Boeing737-500-Kazan2013-scheme.png/320px-Tatarstan-Boeing737-500-Kazan2013-scheme.png
Look familiar?

Earlier, someone made a reference to this article, about how the trimming systems on these aircraft work:

Roger-Wilco | Do you really understand how your trim works? (http://www.roger-wilco.net/do-you-really-understand-how-your-trim-works/)

In that article, towards the end, there is this passage. Note how well it could apply to this crash, and the Kazan crash:

I have watched, in the simulator, a 737 go-around from a Cat lll fail passive approach (as described above) with its marked pitch up; HP kept his arms locked forward to contain the attitude whilst simultaneously running the trim forward with the thumb switch. I am sure he was expecting the trim to reduce push needed and he either didn’t know, or had forgotten, that it wouldn’t. We duly pitched straight back quickly into the ground as the tailplane incidence ‘bit’.

[Chesty Morgan]

De facto, what is your in trim speed if you're out of trim AND? Higher or lower than your current speed?

The object is to minimise attitude change/altitude divergence (especially at low level) before you can rectify whatever problem induced the unusual attitude/out of trim position. If you have a nose low unusual attitude induced by an AND mis-trim and reduce thrust you will exacerbate the situation. Maintain or increase thrust initially, control attitude as well as you can and then deal with the problem.

[xollob]

FGD135 that profile is scarily similar, prompted me to read more about it
Wiki page Tartasan 363 details (https://en.m.wikipedia.org/wiki/Tatarstan_Airlines_Flight_363)

Also interesting is the other bit at the end of understanding your trim article link, I can’t cite with certainty any accident that has been caused by doing this, but I strongly suspect this was a factor in the infamous Icelandair upset event at Oslo The aircraft went quickly from +20 deg to -40 deg and was only saved from a CFIT by a 3.5g pull up, bottoming out at 360ft. Sadly, the report does not discuss the control inputs, nor does it contain any FDR traces, so this trim confusion explanation must remain speculation. I would be astonished, however, if there weren’t more examples of this error, particularly in unfamiliar situations.

[MrSnuggles]

Popping in from lurking, will be gone in just a sec.

Ranger One and Aluminium Shuffler

Talking to the airplane would make sense in a way. I am totally talking to my car sometimes - confusing as it is to my passengers, LOL!

Y'all

What do you make of the strange bump in mommaklee's data? Gusts/drafts/the wrath of Mammon? Some kind of mishandling? Or just FR24 junk data?

[737er]

Not quite.The autopilot trim system also runs at two speeds ( high or low) depending on the flap position.




Well sort of. The high speed AP trim though is equal to the manuel electric trim speed with flaps up. The AP flaps up trim speed slower still.

From Boeing:


Normal electric trimming of the stabilizer is done at one of two rates as controlled by flap position. Trim
rate with flaps retracted is 1/3 the trim rate with flaps extended. The autopilot actuator also trims at
one of two rates as controlled by flap position. High speed autopilot rate is equal to the normal electric
low speed rate. The low speed autopilot rate is 1/2 the rate of the high speed autopilot rate.




So in manuel flight with flaps out , the trim wheel moves three times faster than depicted in the video.

[silvertate]

This accident is strikingly similar to the Tatarstan Airlines Flight 363 B735 that crashed at Kazan, Russia in 2013. This is the graphic of the vertical profile in that accident:

Interesting, not seen that one before.

One possible reason for this profile is something I have seen in light aviation training - a confusion between stalling and zero g. If the push results in reduced g or slightly negative g, it can be percieved as the aircraft stalling, and so pic continues to push forwards to 'unstall' the aircraft. And the result is similar to that profile.

And this is not something that all pilots have trained for, as aerobatics are not on the syllabus. And is not something you can train for in the simulator. And there is a big difference in individual responses to odd attitudes and zero g. Some students shout 'yeeeehaaa' and ask to do it again, while others freeze. And it is not always the bold and confident student who shouts 'yeeehaa'. But if the reaction to zero-g has never been addressed and mitigated back in flight school, it is never going to be so in commercial flying. (Especially with spinning off the syllabus now.)

This is where a post-CPL course is aerobatics and gliding would be beneficial to all new candidates, to stretch their horizons and abilities. Experiencing an upset for the first time with pax on board is not the best time to discover you are uncomfortable with the sensation, or don't have an automatic stall identification and response reflex (Air France please note).

[aox]

Also, for what it's worth, remember the cloud base was in the vicinity of 1700 ft above *MSL* (I am using GPS altitude in these plots).

I had the impression it might be a bit higher than that, though it isn't worth anything.

I've watched three videos of this, one from several miles upwind more or less in line with the runway (posted by skadi at #468).

It looks like the disappearance in cloud is about double the height of the low point before climbing, but that's only an approximation.

The video has the aircraft in cloud for about 34 seconds. Fitting that on your height plot, that would have the base slightly over 2000 feet (edit: amsl)

Incidentally, also about that video, there isn't any apparent sideways movement, so I still haven't figured out why some people are talking about a bank angle.

[fireflybob]

This is where a post-CPL course is aerobatics and gliding would be beneficial to all new candidates, to stretch their horizons and abilities. Experiencing an upset for the first time with pax on board is not the best time to discover you are uncomfortable with the sensation, or don't have an automatic stall identification and response reflex (Air France please note).

Well said, silvertate! Or better still make it part of the CPL course.

In D.P.Davies' classic book written in the 1960s "Handling the Big Jets" the author suggests all line pilots should do 5 hours per year in basic a/c practising recovery from unusual attitudes/positions.

I also agree with your comments about reaction to zero-g.

[Rananim]

The AND trim was either accidental(I agree highly unlikely) or an overcompensation for the pitch attitude associated with a GA with low weight aft cg and sustained gale-force headwinds.
The Dont do this is either from the PF if the trim input was accidental or from the PM if it was overcompensation from PF.
The Jammed elevator sim scenario teaches pilots to respect the power of the THS.Anticipation and the dangers of overtrimming are quickly appreciated when doing this training.Short bursts of trim(esp when flaps down) is key.Concern with extreme ANU/AND attitude can sometimes lead a pilot to overcompensate.
If a runaway (ac problem) then why the Dont do this?

[Raffles S.A.]

Will be interesting to see if the windshear alert sounded and/or stick shaker activated at any time.

[FullWings]

There are parallels being drawn with TA363 and there are some similarities but at the same time, some significant differences.

TA363 was about 80-90kts slower so a push-over was necessary to avoid a stall. The FlyDubai trace shows a segment of high pitch (consistent with encountering a positive shear and trying to keep constant IAS, whether manual or automatic) but a recovery to normal attitudes and IAS afterwards. The aeroplane then followed an unremarkable flightpath for 10-15s until pitching down rapidly towards the ground, accompanied by expressions of incredulity from the pilots.

There is very little that supports any kind of stall. If the stabiliser had stalled with a positive AoA, the end result would have been less of a pitch down moment, although the blanking effects on the elevator are uncertain. There probably was icing around but why should the aircraft suffer serious adverse effects from it when accelerating to the highest IAS for some time? You’d have expected it to happen at slow speed and high AoA, neither of which were present when the LoC appears to have happened.

At the moment it’s pointing much more towards some kind of flight control failure; whether or not it was recoverable we’ll have to wait and see.

[Aerospace101]

There are some significant human performance issues raised in the TA363 report . From Aviation Herald:

The MAK reported that the French BEA performed a sensorical simulation with the aim to model how the crew could lose spatial orientation. The model suggested that after initiating the go around the feel of the pitch angle of the crew and the actual pitch angle were in agreement until about +14 degrees, when the nose rose further to 25 degrees the crew still felt the pitch increase to 17 degrees. Subsequently, when the actual pitch began to decrease the felt pitch continued to increase until reaching +25 degrees, in this phase first nose down inputs are recorded on the flight controls. The felt pitch and actual pitch subsequently completely separated, as the aircraft settled in the dive further increasing nose down inputs were recorded.

The MAK reported that the British AAIB performed simulations with the aim to determine whether somatogravic illusions were present. The AAIB concluded that in the absence of proper control of the instrument readings the crew could have perceived during the transition from climb to dive after the missed approach, that the aircraft was flying inverted.

The MAK performed simulator tests with a number of pilots having them go through a scenario similiar to the accident flight, in particular forcing a go around at intermediate height with the autopilot disconnecting at the initiation of the go-around by pressing the TOGA button. The MAK reported that the vast majority of crews coped well with the scenario but found it difficult to master reporting highly increased stress levels, especially when the pilot monitoring did not provide full assistance. A number of pilots, although the autopilot disconnect aural and visual alerts are very distinct and have high attraction potential, did not catch the fact, that the autopilot had disconnected, several silencing the alerts by pressing the AP disconnect button, a number (about 42% of the pilots tested) not recognizing the alert at all and therefore responding with a substantial delay or not reacting at all. None of the pilots participating in the test was able to answer all questions to the procedures correctly, the MAK reported that 28% even believed the go around was automatic on autopilot despite the AP disconnect alert indicating lack of knowledge and a substantial gap between theoretic knowledge and practical skills.

Of all pilots participating in the test only one third mastered the go around successfully. Only 28% attempted to achieve a suitable pitch angle after initiating the go around aiming for +15 degrees of nose up, others began to react only between +20 and +37 degrees of nose up attitude and airspeeds as low as 90 KIAS with stick shaker activation. None of the pilots was able to level off at the assigned altitude.

In a second part of the experiment a test pilot produced a pitch up upset similiar to the accident flight and then let the participating pilot recover the aircraft. None of the pilots took the right decisions and none was able to recover the aircraft. The most common mistake was to believe, the control wheel would return to the neutral position on its own, this mistake however resulted in a substantial acceleration of the nose down movement resulting in rapid increase of the dive and vertical accelerations between +0.5G and -1.2G. The MAK annotated that in real flight such an acceleration likely causes the temporary and permanent incapacitation of passengers, cabin crew and even flight crew and may cause injuries to occupants.

The common mistakes during the upset recovery noted by the MAK were non-optimal application of flight controls especially if the aircraft is in a bank, no reselection of flaps in order to adjust to the current airspeed resulting in flap limit exceedance and loss of additional height, the non-use of speed brakes. The MAK reported, that after demonstration of the correct upset recovery technics almost all pilots were able to apply the technics and recover the aircraft, suggesting that the result of the experiment was mainly the result of lack of pilot training with respect to upset recovery.

[anson harris]

Total speculation here, and nothing I can point to in support of this, but I'm willing to make a bet that wind shear wasn't a factor.

[Centaurus]

I noticed nothing said about recovering from a extreme nose high attitude and low airspeed by rapidly unloading and rolling the aircraft with full aileron/spoilers into a 45-60 degree angle of bank to get the nose to drop. That is the advice in the 737 FCTM.

Also, to clear the decks as it were, the flight director should be immediately switched off if encountering a gross unusual attitude. This action is to avoid unwanted distraction of FD needles blocking the view of the basic raw data as displayed by the ADI. FD use and subsequent addiction as espoused by many operators can be potentially deadly in extreme UA recoveries

[GNS APPROACH]

Since 2000, several incidents have involved extreme nose-down attitudes during the go-around maneuver on different types of airplanes from different manufacturers. These incidents often result from a breakdown in correct cockpit instrument scanning. Here are some examples:

After the PF initiated a manual go-around at night over the sea, at 1,000 feet (305 meters), the PF kept a prolonged pitch-down input resulting in a 15-degree nose-down attitude and a dive that was not recovered before the impact with the sea. The amplitude and duration of the initial reaction by the PF to the “pull-up” warning from the ground proximity warning system (GPWS) was insuficient (i.e., a full back stick input was required).

After the PF initiated a manual go-around in instrument conditions, and approaching 2,500 feet (762 meters), the ight director altitude capture mode was activated earlier than expected by the crew because of a high rate of climb. The PF manually initiated a level-off but kept a prolonged pitch-down input that resulted in a dive that reached an extreme negative attitude (minus 40 degrees). The PF recovered from the dive at about 400 feet (122 meters) above the ground with a vertical acceleration of 3.6 g-force (g).

[spitfire_sl]

Video of a runway stabiliser in a 737CL simulator.

watch?v=3pPRuFHR1co

[tarik123]

Instead of all the speculations and redesigning the B737. does anyone have any information from the voice or data recorders?
You remind me with the last crash of Air Asia, were a program was made about it assuming that the weather was the cause of the crash, and at the end, the reason was the captain pulling circuit breakers during cruise.

[SLFstu]

@GNS APPROACH
Extreme nose-down attitude

Boeing drivers would no doubt be aware, but for the rest of us, including interested lurkers like me, don’t you think you should have acknowledged the source and context of the 3 paragraphs of your post, that you lifted verbatim from the following Boeing article:

"Performing Safe Go-Around Maneuvers”, Aero Magazine, Issue 55, 3rd Quarter 2014.
Available online at http://www.boeing.com/commercial/aeromagazine/articles/2014_q3/pdf/AERO_2014q3.pdf

Further on in that article is this rather alarming statement (which may or not be relevant to this crash):
“When pilots are distracted, an airplane could change its flight path from a normal go-around climb to a steep dive in fewer than 10 seconds.”

[Machinbird]

Did the aircraft have the Speed Trim System?
Haven't seen any discussion about that, but I would think it would apply.

Speed Trim System
The speed trim system (STS) is a speed stability augmentation system designed to
improve flight characteristics during operations with a low gross weight, aft center
of gravity and high thrust when the autopilot is not engaged. The purpose of the
STS is to return the airplane to a trimmed speed by commanding the stabilizer in
a direction opposite the speed change. The STS monitors inputs of stabilizer
position, thrust lever position, airspeed and vertical speed and then trims the
stabilizer using the autopilot stabilizer trim. As the airplane speed increases or
decreases from the trimmed speed, the stabilizer is commanded in the direction to
return the airplane to the trimmed speed. This increases control column forces to
force the airplane to return to the trimmed speed. As the airplane returns to the
trimmed speed, the STS commanded stabilizer movement is removed.
STS operates most frequently during takeoffs, climb and go-arounds. Conditions
for speed trim operation are listed below:
• Airspeed between 100 KIAS and • Autopilot not engaged
Mach 0.5
• Sensing of trim requirement
• 10 seconds after takeoff
• 5 seconds following release of
trim switches

Offering this for consideration of possible failure modes.

[de facto]

Yes it does.

[Wrist Watch]

Instead of all the speculations and redesigning the B737. does anyone have any information from the voice or Data recorder? The last official statement from the Russian Interstate Aviation Committee was on the 23th of March and it included the following:
The Air Accident Investigation Commission jointly with the experts from the participated States continue work on addressing the possible scenarios of Boeing 737-800 A6-FDN (FlyDubai) aircraft accident.

The meteorological conditions near Rostov aerodrome are being analyzed, as well as aerodrome aids, ATC and AirMet service actions. ATC data from other aircrafts that conducted their flight in the same area are used during this analysis.

In the IAC Laboratory, the work on the flight recorders' and ground aids' data is continued.

With the use of onboard recorders' data and all necessary flight and engineering documentation the aircraft systems' operationability is being analyzed, including the aircraft and engines control systems. The flight crew state during the accident flight is also being analyzed.

The identification of recovered aircraft fragments and the layout is in progress.
The media in their sensationalist manner have published something they believe to have been the 'last words' and they already have the probable cause determined, together with safety recommendations.
As always, it'd be wise to stick with official sources.

[oicur12.again]

“Except perhaps when ALTERNATE LAW comes into play?”

Of course. It goes without saying. But the bus is in normal law 99.9999% of the time and has saved MANY people from screw ups that a 737 cant save you from.

“Yes, because the best solution for pilots loosing basic stick and rudder skills is to just put a computer to do that for them…..”

What is a better solution? Removing technology that provides protection?


“... And we know how that turned out with AF447...”

Yes, a poorly trained crew will always find ways to crash a plane though.

I find it truely alarming that so many pilots on this website have such vastly differing opinions about how FD's work and stab trims and windshear recoveries.

Shouldnt we ALL be precisely on the same page with these basics?

[Rescue3]

Oh, there has been some super analysis and some speculation on the 47 odd pages.
Should we not await the DFDR & CVR details, surely that is the truth...

[Machinbird]

The multiple posts (by at least one) suggesting this crash may have resulted from tailplane icing reminds me of an earlier thread in which many seemed convinced tailplane icing was the likely culprit.
It was Colgan (Buffalo) in February 2009. As I recall, the Colgan crew had been discussing tail plane icing on the way in towards Buffalo. Probably as a result of that discussion, the FO retracted the flaps without a command and without announcing her actions.:eek: Then the CPT compounded it by overriding the stick pusher.

Tail plane icing is most likely to create a stabilizer stall with the flaps down.
We still do not know the flap position on this accident aircraft. We will probably have have to wait a bit to learn from the preliminary accident report.

Using the engine pitch moment to raise the nose is counter-intuitive in a nose low departure.
Unless you already have that trick in your bag of tricks, you are unlikely to use that technique when the chips are down.

[Intruder]

Except that the airplane type involved in the Colgan crash was NOT prone to tailplane icing.

Despite that fact, the airline used a 'canned' icing training video that emphasized recovery from tailplane stalls in OTHER airplane types caused by icing. The Captain had recently gone through that training session, and had the WRONG recovery procedure impressed in his mind.

[Chronus]

The go around was initiated fairly early, what could have been the reason, other than wx.
I`d suggest the start of the causal sequence chain was in the intermediate approach phase. The crew reacted by throwing away the landing and taking the problem back up to resolve it. If indeed the issue was stab trim related then flap down selection would have exacerbated it by the ensuing pitch down, which with a locked nose down pitch trim and the high energy of the go around and the normal crew reaction of reducing thrust, with an underwing engined aircraft, would also add the nose down pitch angle.
In such a scenario I cannot see how fault and error may be attributable to crew error.

[RatherBeFlying]

SLFstu's link to Performing Safe Goarounds:
http://www.boeing.com/commercial/aeromagazine/articles/2014_q3/pdf/AERO_2014q3.pdf

includes a short basic tutorial on instrument scan.

Interestingly it comments that acceleration while pitched down creates the illusion of a climb:uhoh:

The article describes a number of pitch down incidents where the crew recovered at low altitude and one that did not recover in time.

[Katzenjammer kid]

Video of a runway stabiliser in a 737CL simulator

Thanks for the training video, it is very worrying that they don't know the difference between the elevator and the rudder.

[Ian W]

But the bus is in normal law 99.9999% of the time and has saved MANY people from screw ups that a 737 cant save you from.

“Yes, because the best solution for pilots loosing basic stick and rudder skills is to just put a computer to do that for them…..”

What is a better solution? Removing technology that provides protection?


Well to follow your argument to a conclusion which is supported by numerous bean counters in your airline and others, the best cost effective solution is to remove the pilots. Had the subject aircraft been controlled by automatics only, it would probably not have crashed; no fatigue issues leading to 'pilot error'. Had AF447 had software that did not have an 'otherwise case' that said give it to the crew - the aircraft may not have crashed. In many recent cases the 'wet ware' in the cockpit has been the problem mishandling the automatic systems. Airbus with its 'human-on-the-loop' approach is easiest to convert to 'optionally manned'.

Be careful with your arguments you may end up as employable as elevator operators.

[Milvus Milvus]

Interesting video of the stabiliser screwjack. What position would the stabiliser end up in if that screw jack fractured ?

[gums]

Salute!

How come no one here has looked back at the Norwegian incident with the 737?

ASN News » Report: Frozen elevator nearly puts Boeing 737-800 into a stall (http://news.aviation-safety.net/2015/03/30/report-frozen-elevator-nearly-puts-boeing-737-800-into-a-stall/)

There is more than one way to have frozen stuff kill you. Once even had ice crystals in my hydraulics that kept a main gear up!

worth thinking about...

[Centaurus]

Video of a runway stabiliser in a 737CL simulator.


Couple of amateurs in that cockpit especially the bloke doing all the talking in the left hand seat. The stabiliser trim handles should have been used to obtain rapid operation manually. That is why the handles are there.

Instead the LH seat operator wasted what could be critical time making feeble hand movements on the top of the stab trim wheel like a PPL and his Cessna. If that is supposed to be an official published "training" film then the Regulatory authority needs to audit the series published by the film makers. :=

[ZFT]

As the same Regulatory Authority approves their ATO for Type Ratings across 10 different aircraft types utilising 65 different FSTDs located across 20 facilities in 16 different countries, I wouldn't hold out much for competent audits!!!

[ManaAdaSystem]

Salute!

How come no one here has looked back at the Norwegian incident with the 737?


How come you didn't bother to read the thread before you postet?

[Sunamer]

Sorry but this vid seems not to be real, it is just physically impossible IMO
That is because that video specifically states "Genuine Boeing 737-300 series throttle quadrant converted for flight simulator use. "

What makes you think that it was a certified flight sim, and not some home-based FSX/Prepar3d toy?

[ZFT]

Sunamer,

What makes you think that it was a certified flight sim, and not some home-based FSX/Prepar3d toy?

Flight Simulators (apart from BITD) are never certified, they are always uniquely qualified.

[Kulverstukas]

Kommersant newspaper published article about "pilot disagreement which led to disaster (http://www.kommersant.ru/doc/2949365)" today. It sounds crazy.

[c_coder]

https://translate.google.com.au/translate?hl=en&sl=ru&tl=en&u=http%3A%2F%2Fwww.kommersant.ru%2Fdoc%2F2949365

[DaveReidUK]

Somewhat confusing article:

Russia May Suspend Boeing 737 Airworthiness Certificates (http://sputniknews.com/russia/20160328/1037064364/russia-may-suspend-boeing-flights.html)

[737er]

Fantastic article on pitch trim and I believe scenario #1 is a distinct possibility.

http://www.skybrary.aero/bookshelf/books/2627.pdf

[klintE]

"According to preliminary data, they quarreled, and at the same time pulling the steering wheels in opposite directions, which is why passenger plane went into a dive and crashed."
from:http://www.dni.ru/society/2016/3/28/332211.html
So, we got some kind of 'dual input' here. Unbelievable.

[Alientali]

[I]"According to preliminary data, they quarreled, and at the same time pulling the steering wheels in opposite directions, which is why passenger plane went into a dive and crashed."

So, we got some kind of 'dual input' here. Unbelievable.

Just reading this... apparently, they don't see the difference between stick and elevators. Various media outlets keep quoting each other, repeating each other's errors and adding their own.

More quotes, from http://www.kommersant.ru/doc/2949365:

"Its elevator [I think they meant "stick" here] was set for a dive, while the stabilizer was exerting an opposing force, making the plane's nose point up". (???)

"The plane's computer began to receive electric signals from all actions performed by pilots on both yokes; normally the yokes work synchronously."

"In the previous landing attempts, autothrottle wasn't working well."

And they point to high AoA on GA and stall as the cause of the crash.

It's a bad, bad, bad case of garbled transmission - I assume someone close to investigation *is* informing the media, but is explaining the findings in layman's terms. The media, in turn, thinks those aren't layman's enough for the general public... and in total absence of technical knowledge, kills it completely.

[six7driver]

737 er it is a great article, if the airplane is being hand flown this combined with a disorienting somatogravic illusion would test any pilot to their limits, at 4am in the morning after a long duty day this would be beyond most of us surely.

[KRH270/12]

Sorry, but why do you speculate about tailplane icing and, stab out of trim?

The Plane impacted with massive bank angle!

Pitch authority problems alone almost allways result in nearly wings level CFIT.

[MrSnuggles]

Kulverstukas

You did an awesome job with the MH17 reporting.

May we borrow your time for some equally awesome reporting this time?

[Bonzo777]

https://www.rt.com/news/337400-pilots-conflict-boeing-rostov/

[Sunamer]

they are always uniquely qualified.
Semantics really...
What makes you think that that particular ehh...sim was "uniquely qualified"...
especially, given that in the video that trim wheel clearly doesn't have enough speed to move the stab to the position the stab indicator displays at the end of the video.

[aox]

The Plane impacted with massive bank angle!

Some people keep saying that, but on what basis?

A video from some distance upwind and in line with the approach shows no sideways motion at all.

[ZFT]

Semantics really...
What makes you think that that particular ehh...sim was "uniquely qualified"...
especially, given that in the video that trim wheel clearly doesn't have enough speed to move the stab to the position the stab indicator displays at the end of the video.

Err I didn't think anything of the kind and no not semantics if you have any comprehension of the difference.

[KRH270/12]

Some people keep saying that, but on what basis?

A video from some distance upwind and in line with the approach shows no sideways motion at all.


The videos show large bank angle.

People always assume bank = turn , but thats not always true, especially not for very high bank angles close to 90 degree with massive pitch down. Thats why a chrash on centerline does not mean no bank!

[Capt. Inop]

Fantastic article on pitch trim and I believe scenario #1 is a distinct possibility.

http://www.skybrary.aero/bookshelf/books/2627.pdf


Exactly what I suspect is the case in this accident.
Too many Boeing pilots don't know how the trim system on the plane they fly works, especially those who come from other types with different trim system.

[enola-gay]

The video of the Bagram crash shows a B747 stall in the climb and immediately roll 90 deg without turning, then goes nose down into terrain. A high bank angle with stalled wings produces no turning moment.

https://www.youtube.com/watch?v=_VyHa05LyIA

[Milvus Milvus]

https://www.rt.com/news/kazan-plane-crash-video-882/

[Capn Bloggs]

Fantastic article on pitch trim and I believe scenario #1 is a distinct possibility.

http://www.skybrary.aero/bookshelf/books/2627.pdf
That was interesting. I was always taught to not fly with the trim; if the speed didn't change, the trim wouldn't change. You may have to pull and push to move the nose around, but the trim wouldn't change. With underslung engines (and no FBW to compensate as the power changes), though, how much trim would you need to compensate for say GA thrust? Is it a lot? Is the ideal procedure to hold some/a lot of forward pressure on the control column until the thrust comes off as you level off? How does the autopilot cope with a GA thrust climb?

[fireflybob]

I've been following this thread from the start. Correct me if I am wrong but this was two Captains flying together?

An early CRM course I attended showed evidence that Captains flying together were statistically more likely to be involved in an incident/accident. Also the data showed that two Training Captains flying together was an even worse crew combination.

Something to do with the cross cockpit command authority gradient.

[striker26]

https://www.rt.com/news/337400-pilots-conflict-boeing-rostov/

If this report is true (not fully believing it until the official reports come out) it seems the Captain wanted to counter act the windshear/turbulence by climbing aggressively with full power while performing the GA (apparently pulled back too much) . The climb rate was so great (light a/c, flaps retracted) , the FO counter acted the Captain's decision in disagreement and the a/c was given conflicting commands and stalled. Seems like the Captain wanted to make a 3rd attempt manually and was frustrated (anyone would be)...question is did they attempt stall recovery and failed with TOGA thrust set the whole way after they realized the change in g-force? Pretty scary situation...Any thoughts guys?

[Otto Throttle]

Fireflybob

It was a standard flight crew complement. The FO may have had command experience previously (very likely given total time), but he was not a FZ commander.

[fireflybob]

It was a standard flight crew complement. The FO may have had command experience previously (very likely given total time), but he was not a FZ commander.

Otto thanks for the clarification

[Milvus Milvus]

RT is suggesting opposite inputs from the two pilots separated the control column .....

"When one of the control columns becomes jammed, the crew can override (breakout) the faulty control. The control columns are interconnected below the cockpit floor by a torque tube with a device that enables the controls to be separated from each other. The Elevator Breakout Mechanism connects both control columns by two springs which will separate the columns when ± 30Lbf/13Kgf is used to overcome them. When applied, the control columns are mechanically separated from each other. Note that deflection of the elevators is significantly reduced and a higher force is needed to move the elevators. (even higher than with manual reversion)"

[CaptainMongo]

https://www.rt.com/news/337400-pilots-conflict-boeing-rostov/


With the caveat this reporting could be completely wrong:

The article indicted the crew attempted two auto lands where earlier in this thread it was stated only manual landing with use of a the hud were the only authorized company specs. (Personally I think this is bad reporting)

"Meanwhile, according to Kommersant’s sources, the recorder revealed that flight FZ981 had made two attempts to land in automatic mode, and since a veering squall wind hampered the second attempt, the pilot decided to make a third approach for landing in manual mode."


If the aircraft was in windshear where the GPWS recognized it, the warning would have clearly been heard on the recording and I would have thought that would have been reported.

If the aircraft was not in actual windshear a normal go around should have been accomplished (to maintain energy), only reverting to the windshear recovery procedure when the pilots observed or GPWS reported Windshear.

I find the possible dual control situation extraordinarily troubling.

[de facto]

During,manual go arounds,never needed much trim if any....i would believe if they needed a high amount of trim nose down would be if they initially way over pitched the aircraft.(whatever the cause may be).

[FullWings]

If this report is true (not fully believing it until the official reports come out) it seems the Captain wanted to counter act the windshear/turbulence by climbing aggressively with full power while performing the GA (apparently pulled back too much) . The climb rate was so great (light a/c, flaps retracted) , the FO counter acted the Captain's decision in disagreement and the a/c was given conflicting commands and stalled. Seems like the Captain wanted to make a 3rd attempt manually and was frustrated (anyone would be)...question is did they attempt stall recovery and failed with TOGA thrust set the whole way after they realized the change in g-force? Pretty scary situation...Any thoughts guys?
Yes. Don’t buy a used car from a Russian news agency.

Unless the FR24 data is very wrong, it shows a loss of control but it is difficult to see a stall in there, in fact quite the opposite. For reasons yet to be revealed it seems there was a pitch control issue, be it self-induced or system failure related.

The reporters for this TV network look like they are in a kind of Russian-English-Russian feedback loop involving this site and random bits of aeroplane manuals. I think we’ve exhausted most avenues of speculation and are awaiting the FDR output...

[Loose rivets]

Was that SIM ride supposed to be taken seriously?


Instead the LH seat operator wasted what could be critical time making feeble hand movements

While the PF seemed to be waggling the stick about perhaps to liven up the smooth air.

And as for that screw jack. I've got antique furniture that sounds better than that when it reached the top of the run.

I've never understood Boeing and that million-turn philosophy. The BAC 1-11 wheel was supremely accurate and quite simply part of the hand flying. It's hard to see why there would be two such disparate design philosophies.

[de facto]

Was that SIM ride supposed to be taken seriously?


Cant be...a real show of what not to do in all aspects of PF/PNF demonstrated by both pilots,especially the guy in the left.
It hurts the eyes for sure.

[Kulverstukas]

You did an awesome job with the MH17 reporting.

Not me. You mixed me with somebody. At least not about MH17.

About this Kommersant' article (and it's direct translation on RT with only their own addition about steering separation) it for me looks not more believable than that "knuppel" business.

MAK issued press-release concerning this latest "investigation results". In short it can be translated into "who are all this people and what are they talking about?".

[Organfreak]

It seems to me that according to the previously-posted article prepared by Boeing http://www.boeing.com/commercial/aer...ERO_2014q3.pdf shows several ways to crash the airplane without any mechanical/technical problem at all! I could be full of it, since I'm not a pilot.

[Lonewolf_50]

It seems to me that according to the previously-posted article prepared by Boeing http://www.boeing.com/commercial/aer...ERO_2014q3.pdf shows several ways to crash the airplane without any mechanical/technical problem at all! I could be full of it, since I'm not a pilot. The link is broken.

[Organfreak]

The link is broken.

So sorry, yes it is. It may have been removed. Checking.

[DaveReidUK]

The link is broken.

http://www.boeing.com/commercial/aeromagazine/articles/2014_q3/pdf/AERO_2014q3.pdf

pp13-17: Performing Safe Go-Around Maneuvers

[Organfreak]

OK, this one works, at least for me:
http://www.boeing.com/commercial/aeromagazine/articles/2014_q3/pdf/AERO_2014q3.pdf

I hope that every pilot who's flying passengers is thoroughly familiar with this information. Some of the responses here convince me that not everyone is familiar with this information. The article is on Page 13.

[Alientali]

From MAK/IAC (Artur Muradyan, aide to chairman):

"At this time, we can neither confirm nor deny the reports of the pilots' last conversation. Neither MAK nor any of its employees gave any information to the media - not to Kommersant, not to anyone at all. When we release our data, it will be open to comment. We haven't yet completed the decoding of CVR and FDR, it's in progress, we're still analysing and compiling the data and identifying the [pilots'] voices. It's possible that we will release some preliminary data in a couple of weeks, but the final results of decoding will be published in a month."

[bsieker]

[...] it seems the Captain wanted to counter act the windshear/turbulence by climbing aggressively [...], the FO counter acted the Captain's decision in disagreement and the a/c was given conflicting commands and stalled.

Talking about a fully mechanically controlled airplane this makes no sense at all. There's a direct mechanical connection between the flight controls and the control surfaces (normally hydraulic, but manual fallback) in the 737. There are no "conflicting commands".

And while a "split elevator" condition arising from both pilots pushing/pulling on the yokes against each other may occur in the 767 (see the final report for Egypt Air Flight 990 (http://www.ntsb.gov/investigations/AccidentReports/Reports/AAB0201.pdf)), I have never heard of this for the 737, and that alone would not automatically cause loss of control and crash, much less a stall.


Bernd


EDIT: Egypt Air 990 was a 767. Thanks, Airbubba.

[olasek]

Talking about a fully mechanically controlled airplane this makes no sense at all.
Makes a lot of sense.
In a "mechanically" controlled 737 (as you said it) dual input is very much possible. I suggest you update your knowledge - it is possible in ANY Boeing. If actions of both pilots are sufficiently different and forces exceed some limits - both sides will be separated, this was done to counter possible jamming of controls on one side, certification required such a solution.

[Aluminium shuffler]

The controls are not "normally hydraulic, but manual fallback" - the pitch controls and ailerons are cable and pulley systems, with hydraulic assistance. Manual reversion simply means the hydraulic (or electrical for stab trim) assistance is removed. The controls can split between the two columns, the captains column connected to left elevator and ailerons, the fo's connected to right elevator and spoilers. They split in case one jams, as Olasek said.

Amazing how many people are commenting here without proper knowledge.

[silvertate]

The controls can split between the two columns, the captains column connected to left elevator and ailerons, the fo's connected to right elevator and spoilers. They split in case one jams, as Olasek said.

Ever tried that in the sim? The system simply does not work, and so nobody has ever tried it to see what happens ( I have asked many times, with no success).

Come on chaps - how many have done a control disconnect in the sim? Looks like another 737 system that does not work.

[Aluminium shuffler]

Yes, once, a long time ago (Classic, but the system is the same as the NG). The breakout force is fairly large, but it works.

[Mr Optimistic]

SLF. Would a 'split' elevator explain the roll/ bank?

[fireflybob]

Seems like a common thread to quite a few accidents in the last few years is inattention to the correct pitch attitudes. (Think of the initial autopilot disconnect on AF447). In the older types the flight director was a fairly crude device with limited modes so it was taken for granted that you needed to know the raw data attitudes for a given phase of flight and also to monitor carefully that these were being flown correctly.

[bsieker]

The controls are not "normally hydraulic, but manual fallback" - the pitch controls and ailerons are cable and pulley systems, with hydraulic assistance. Manual reversion simply means the hydraulic (or electrical for stab trim) assistance is removed.
I know how it actually works (hydraulically augmented cable-and-pulley), but that's not how Boeing describes it in the (admittedly not necessarily technically precise) FCOM:

The pitch control surfaces consist of hydraulically powered elevators [...]
[...]
With loss of hydraulic system A and B the elevators can be mechanically positioned by forward or aft movement of the pilots’ control columns.

So I chose to go with Boeing's description, knowing that it is slightly misleading.

And what I mean is that it makes not sense to say any single entity receives "conflicting commands", since after a split there is no longer a single system, but rather two separate ones (if that is indeed the case; see below). And an airplane as such cannot receivce "conflicting" "commands", it just experiences forces and moments and reacts according to the laws of physics.

The controls can split between the two columns, the captains column connected to left elevator and ailerons, the fo's connected to right elevator and spoilers. They split in case one jams, as Olasek said.

I know (and mentioned in my previous post) that what you say is true in the 767 (as evidenced by Egypt Air 990), but the description in the manual does not imply this specifically for the 737.

In fact it says that both elevators are linked by a torque tube, which is mentioned separately from the so-called "override mechanism", which allows the two control columns to be separated in the event that one of them becomes jammed. This does not really imply that it is the torque tube which is broken to free a jammed control column, but rather the connection between the two control columns.

Amazing how many people are commenting here without proper knowledge.

Instead of bemoaning people's lack of knowledge you could just respond to my wondering if the 737 systems are indeed supposed to work the same as in the 767.

If my understanding of 737 is demonstrably wrong, please correct me by citing authoritative sources (which I agree the FCOM is not for a technical description; but it often gives a good idea.)


Bernd


EDIT: Egypt Air 990 was a 767, Thanks to Airbubba for the correction.

[Chesty Morgan]

SLF. Would a 'split' elevator explain the roll/ bank?
Probably not to the extent postulated in that short amount of time. If unchecked it would eventually but the elevators have a very small moment arm.

[Superpilot]

https://www.youtube.com/watch?v=slL2KMoM3p8

[hec7or]

Elevator Control Column Override Mechanism

In the event of a control column jam, an override mechanism allows the control columns to be physically separated. Applying force against the jam will breakout either the Captain’s or First Officer’s control column. Whichever column moves freely after the breakout can provide adequate elevator control. Although total available elevator travel is significantly reduced, there is sufficient elevator travel available for landing flare. Column forces are higher and exceed those experienced during manual reversion. If the jam exists during the landing phase, higher forces are required to generate sufficient elevator control to flare for landing. Stabilizer trim is available to counteract the sustained control column force.


there's no individual left/right elevator split

......also of interest and applicable to the NG, but not the 300-500 Efis

During high AOA operations, the Stall Management/Yaw Damper (SMYD) reduces yaw damper commanded rudder movement. The EFS module increases hydraulic system A pressure to the elevator feel and centering unit during a stall. This increases forward control column force to approximately four times normal feel pressure. The EFS module is armed whenever an inhibit condition is not present. Inhibit conditions are: on the ground, radio altitude less than 100 feet and autopilot engaged. However, if EFS is active when descending through 100 feet RA, it remains active until AOA is reduced below approximately stickshaker threshold. There are no flight deck indications that the system is properly armed or activated.

[bsieker]

there's no individual left/right elevator split

Thank you. That was my reading as well.


Bernd

[Mr Optimistic]

CM. OK, thanks.

[porterhouse]

If my understanding of 737 is demonstrably wrong,
Yes, it is wrong. But I am afraid you will believe whatever you want to believe. 737 is no different than 757 or any other Boeing.

[bsieker]

Yes, it is wrong. But I am afraid you will believe whatever you want to believe. 737 is no different than 757 or any other Boeing.

Sorry to say so, but your say-so is not what I meant by "authoritative source". It would require some other document contradicting the 737 FCOM and mine and hec7or's reading of it. For example some documentation explaining that the torque tube between the two elevator surfaces is frangible and supposed to be broken by the override mechanism for dealing with a jammed control column.

The 757's flight controls are in fact quite different from the 737's, e. g.

[...] there is no manual reversion.

which already makes it fundamentally different.

It also has three hydraulic systems instead of two.


EDIT: Thanks, Airbubba, Egypt Air 990 was a 767, but everything I said also applies to the 767, including the quote from the FCOM.

Bernd

[Airbubba]

And while a "split elevator" condition arising from both pilots pushing/pulling on the yokes against each other may occur in the 757 (see the final report for Egypt Air Flight 990)

Instead of bemoaning people's lack of knowledge you could just respond to my wondering if the 737 systems are indeed supposed to work the same as in the 757.


Just for the record, Egyptair 990 was a B-767-300ER, not a 757.

A lot of 757 and 767 systems are similar but many, especially hydraulics, are significantly different. The two planes were literally designed by two different groups at the same company. Or, so they tell me...

[bsieker]

Just for the record, Egyptair 990 was a B-767-300ER, not a 757.
So it was.

Thanks for the correction.

[Capn Bloggs]

During,manual go arounds,never needed much trim if any....i would believe if they needed a high amount of trim nose down would be if they initially way over pitched the aircraft.(whatever the cause may be).
Thanks Defacto.

[FDMII]

I don't know why but, FWIW, the Boeing FCOMs in general are notoriously poor by comparison with other manufacturers' FCOMs in providing aircraft technical information for flight crews, so misconceptions are bound to occur. For some here, a bit of professional courtesy and patience is necessary while others are sorting the question out. Airbus, Embraer & Dornier are far more detailed, helpful and well illustrated.

[aterpster]

FDMII:

I don't know why but, FWIW, the Boeing FCOMs in general are notoriously poor by comparison with other manufacturers' FCOMs in providing aircraft technical information for flight crews, so misconceptions are bound to occur.

Then again, major airlines, at least in the US., do not (or did not during my tour of duty) use Boeing's or other OEMs FCOMs. They wrote their own, tailored to their training methods, and with FAA concurrence.

The concept worked great at TWA. TWA called them FHBs, for Flight Operating Handbook. So, whether it was a 707, 727, 767, DC-9, MD-8x, L1011, all the FHBs had "that familiar appearance."

[barit1]

Airbubba:
A lot of 757 and 767 systems are similar but many, especially hydraulics, are significantly different. The two planes were literally designed by two different groups at the same company. Or, so they tell me...

Boeing has used tandem design teams for decades. The B-29 and B-50 e.g. The two teams "leapfrogged" each other, with only general exterior similarity. In the case of 757/767, they coordinated on general cockpit design and handling qualities, but little else.

[CONSO]

In the case of 757/767, they coordinated on general cockpit design and handling qualities, but little else.

True- what few realize is that the 757 was largely a Renton design team , and the 767 was largely a Everett design team. Yes there was some interplay on common items and a lot of control/backup/baseline coordination mainly based on what worked historically. The 737 was largely a Renton design team then 747 was largely an Everett design team. And there were some employee transfers between them.

The 777 was mostly a everett design team. The 787 was mostly Everett but had more than a few B2 experienced people, etc

[FDMII]

aterpster, concur; same with my carrier. But for those stuck with the Boeing product, a lot was missing.

[rainbow gravity]

And while a "split elevator" condition arising from both pilots pushing/pulling on the yokes against each other may occur in the 767 (see the final report for Egypt Air Flight 990 (http://www.ntsb.gov/investigations/AccidentReports/Reports/AAB0201.pdf)), I have never heard of this for the 737. Sorry to say so, but your say-so is not what I meant by "authoritative source". It would require some other document contradicting the 737 FCOM and mine and hec7or's reading of it. For example some documentation explaining that the torque tube between the two elevator surfaces is frangible and supposed to be broken by the override mechanism for dealing with a jammed control column.

Here (http://archive.is/739hx) is a good explanation of the B737 Aircraft Control Override Devices (in agreement with Olasek in #969 (http://www.pprune.org/rumours-news/576325-b-738-crash-russia-rostov-don-post9326175.html#post9326175)):

« B737 Aircraft Control Override Devices: There are two devices that allow you to control the aircraft in case of a malfunctioning or jammed control system.
One concerns roll control. When one of the yoke cables (or aileron PCU/spoilers) becomes jammed or moves freely, the opposite control is still available to roll the aircraft. The two yokes are interconnected at the base of the co-pilots control column by the Aileron Transfer Mechanism through torsion spring friction and a “lost motion device”. If the FO control jams, the spring force can be overcome by the Captain thereby controlling the aileron PCU through cables. If the Captain control jams, the FO can control roll by use of the flight spoilers. Note that this only happens when the yoke has been turned ± 12° which engages a so called “lost motion device” which in turn operates the flight spoilers.
The second is related to pitch control. When one of the control columns becomes jammed, the crew can override (breakout) the failing control. The control columns are interconnected below the cockpit floor by a torque tube with a device that enables the controls to be separated from each other. The Elevator Breakout Mechanism connects both control columns by two springs which will separate the columns when ± 30Lbf/13Kgf is used to overcome them. When applied, the control columns are mechanically separated from each other. Note that deflection of the elevators is significantly reduced and a higher force is needed to move the elevators. (even higher than with manual reversion) »

[AerocatS2A]

Just curious: If the 737 elevator can't be split then what happens if the jam occurs in the elevator itself?

[darkwing]

MOSCOW - Russian state television has aired the final words, of the pilots of the flydubai passenger jet that crashed in Rostov-on-Don last week killing 62 people, suggesting pilot error could be to blame.

Russia's Rossiya-1 channel said late on Friday it had obtained a transcript of the last words exchanged a minute before the Boeing 737 plunged to the ground from a source in the investigative commission with access to the plane's voice recorders.

The plane, which was arriving from Dubai, exploded into a fireball last Saturday after missing the runway in the city in southern Russia while making a second attempt to land in heavy wind and rain.

It was carrying 55 passengers and seven crew members. There were no survivors.

PICTURE: Russian Emergency Ministry rescuers search the wreckage of a crashed airplane at the Rostov-on-Don airport on March 20, 2016. CREDIT: AFP

The transcript suggests that the pilot lost control of the plane immediately after switching off the autopilot.

"Don't worry," the pilot says repeatedly in the transcript translated into Russian, before repeating seconds later "Don't do that!" The last words are repeated calls to "Pull up!"

For the last six seconds of the plane's dive, all that can be heard are "inhuman screams," the channel's source said.

The channel suggested -- citing specialists while stressing that this was not the official version -- that the pilot accidentally switched on a stabilising fin at the tail as he tried to pull the plane back to a horizontal position.

With this fin activated, "the plane practically does not react to the pilot's control panel" the channel said.

"The pilots clearly did not understand that the stabiliser was to blame for the steep dive."

The channel said the pilot could have accidentally switched on the button activating the fin due to his reported "chronic fatigue."

The only other possible explanation would be an "unprecedented glitch in the plane's automatic systems," the channel said.

Aircraft control out of sync

Sputnik media, meanwhile, reported on Monday that the Boeing 737-800 could have crashed due to a conflict in the cockpit.

PICTURE: File image of the interior of a Boeing 737-800. CREDIT: Flickr.com/Frans Zwart


According to a source of Russia’s Kommersant newspaper, decoding of the parametric data and voice recorders of the plane revealed that pilots had twice attempted to land the aircraft in automatic mode, but failed to do so due to difficult weather conditions. The pilots eventually decided to come in to land in manual mode.

Experts claim that the crew failed to take characteristics of Boeing 737-800 into consideration, particularly transition from landing to ascend. The pilot on the controls continued to ascend and attempted to pick up speed, while his co-pilot tried to stop gaining altitude. As a result, control of the aircraft was out of sync.

CREDIT: Faisal Akram from Dhaka, Bangladesh, via Wikimedia Commons

The investigation has not yet managed to establish which pilot made a fatal error.

Investigators have launched a criminal probe into whether poor weather, pilot error or a technical fault were behind the crash.

[silvertate]

Just curious: If the 737 elevator can't be split then what happens if the jam occurs in the elevator itself?

Then you would be stuck. It is simply not a good system, and would not pass modern certification.

As I understand the system, you are simply overriding a jammed control column (an unlikely scenario). So any jam in the cabling or in one elevator has no possibility of an override. And even when you do override the other cc (if a cc can jam...), the forces are so high and the control so limited, the aircraft is all but uncontrollable. (Only what training dept said, they have never let me try it after 20 years on the 737.) While other aircraft like the 146 have a complete split between the entire system, with a separated control system to a separated elevator. A much more sensible system, that we could demonstrate on the sim each and every year.

In the case of one pilot pushing and one pulling, on a 737 with a split torque tube, I think the result will be no elevator displacement. The one will cancel the other.





The channel suggested -- citing specialists while stressing that this was not the official version -- that the pilot accidentally switched on a stabilising fin at the tail as he tried to pull the plane back to a horizontal position.

I think this means that he had his finger stuck on the trim-switch, by accident. Trimming in the wrong direction, perhaps, or forgetting to stop trimming. Not sure whay that should happen, but it sort of makes sense of the report here.

[AerocatS2A]

Then you would be stuck. It is simply not a good system, and would not pass modern certification.


This seems very strange. Would be good if someone had a maintenance manual that could confirm this is indeed how it works.

[Aluminium shuffler]

Bsieker, I did not claim that the elevators split, I said the Capt column is connected to the left elevator and the Fo's to the right. That is to get around any jam in the control runs, not just the column. The torque tube between the elevators will move both together, regardless of which column is used. The cables have springs to allow the surface to move off one input while the other is jammed.

As for what happens if the elevator jams completely, silvertate, you fly it on the stabiliser and thrust. It isn't easy, but it should be completely manageable for any commercial pilot. I have done that too in the NG sim, with a simulated impossible, or at least incredibly unlikely, total jam of all roll and elevator control (both columns) - the aircraft is quite controllable on stab trim and rudder. I even had a go (for play, not assessed) at a loss of all flight controls in the box, not being allowed to touch any flight controls from 400 feet on take off. It was perfectly possible to fly the aircraft on thrust alone (no trim, flap or rudder as well as column), like Sioux City - I was given calm and CAVOK, so didn't have to contend with any weather, but I was surprised at how easy it was to land on the runway centreline, in the touchdown zone, and stop it on the runway on the first (only) attempt. The FO managed to do a controlled landing on the runway too, when he had a go at the same.

Furthermore, the 737Max has the same flight control system, the sole difference being the mechanical aileron - spoiler mixer unit being replaced by a flybywire unit that replicates the mechanical one with less weight. So, the system does meet modern certification requirements and in my opinion has more redundancy than any other airliner in current service.

[WYOMINGPILOT]

Bsieker, I did not claim that the elevators split, I said the Capt column is connected to the left elevator and the Fo's to the right. That is to get around any jam in the control runs, not just the column. The torque tube between the elevators will move both together, regardless of which column is used. The cables have springs to allow the surface to move off one input while the other is jammed.

You have said this twice but I have no reference for this so-called Capt. column is connected to the left and FO to the right elevator?


In the event of a control column jam, an override mechanism allows the control
columns to be physically separated. Applying force against the jam will breakout
either the Captain’s or First Officer’s control column. Whichever column moves
freely after the breakout can provide adequate elevator control.
Although total available elevator travel is significantly reduced, there is sufficient
elevator travel available for landing flare. Column forces are higher and exceed
those experienced during manual reversion. If the jam exists during the landing
phase, higher forces are required to generate sufficient elevator control to flare for
landing. Stabilizer trim is available tocounteract the sustained control column force.

u

[sAx_R54]

@Organfreak

Some of the responses here convince me that not everyone is familiar with this information.

Often as disturbing as the initial CFIT in pprune post incident analyses!

[bsieker]

Bsieker, I did not claim that the elevators split

Since I specifically commented on split elevator, It seemed straightforward to take your's and olasek's comments to refer to the elevators.

Thanks for clearing up that misunderstanding.


Bernd

[fab777]

I could refer to an incident that happened in my airline a couple of years ago to raise another scenario that could lead to a loss of control/CFIT. It did happen on a A319, but about the same thing may also happen on a Boeing.

A go-around was initiated at an altitude below the go-around target altitude, but not that much, and the thrust levers were not set to TOGA. Guess what happened when the AP was re-engaged, with the FD still in approach mode? The aircraft dived, still in IMC, to get back to the glideslope, and control was resumed only when they broke out of the clouds. As far as I remember, a disaster was avoided by a couple hundred feet.

Same thing would happen on a B737 if a go-around (or winshear escape maneuver) is conducted without pushing the TOGA switch (and the fact that the gear was still down at impact, see the videos showing the landing and taxi lights, makes me think that was a WS escape maneuver). And then, it is easy to get confused...

Just another wild guess, based on history...

[TURIN]

737aileronFCS.png Photo by patrikcertain | Photobucket (http://s307.photobucket.com/user/patrikcertain/media/737aileronFCS.png.html)


http://www.smartcockpit.com/docs/B737-Flight_Controls_Systems_Summary.pdf

[Centaurus]

The media report said one of the crew disengaged the autopilot. Depending on the severity of the turbulence it is quite possible the autopilot could not cope and in that case it disengages without pilot action. Even if one of the pilots immediately attempted to re-engage the AP, any manually applied control wheel pressures (which could occur in turbulence) will prevent the AP from engaging. If that happens in IMC then recovery success (or otherwise) is entirely dependant on the instrument flying ability of the pilot flying. That, as they say, is another story altogether...

[de facto]

Originally Posted by AerocatS2A View Post
Just curious: If the 737 elevator can't be split then what happens if the jam occurs in the elevator itself?
Then you would be stuck. It is simply not a good system, and would not pass modern certification.


Quote:
Originally Posted by silvertate View Post
Then you would be stuck. It is simply not a good system, and would not pass modern certification.
This seems very strange. Would be good if someone had a maintenance manual that could confirm this is indeed how it works.

Silverstate,

If the elevators are jammed and pilots cant overpower them to clear the jam, wouldnt the stabilizer trim sufficient to fly the aircraft?
I am not sure i understand your reasoning.
NB, for the old armshair SLFs in here,it is a question to make a point,not that i dont know...before you start shivering,crying,once again,oh my god pilots dont know how to fly their aircraft:E

[FGD135]

Hi again all,

I have just made a posting to the Tech Log forum about the behaviour of pitch trim systems in aircraft such as the B737.

I have made reference to this accident, and the Tatarstan Airlines B735 that crashed in Kazan, Russia in 2013. These two accidents have almost identical vertical profiles. Both crashed onto the aerodrome, beside the runway, steeply nose down and at high speed.

You can find it here:

http://www.pprune.org/tech-log/576817-pitch-trim-poorly-understood.html

[Sciolistes]

Can the black box data determine what caused the A/P to disengage or does it just see that the A/P disengaged? Can the FDR data log that the trim switches were pressed or just that the trim was operating? If I were to hazard a guess I would say the FDR cannot determine how the A/P was disconnected only that it was. Which would make leaked comments worthless.

[ipilot86]

If I am correct, there was a recent computer modification available to 737NG. One of the changes is that unsuccessful attempt to engage the autopilot would cause the same indication (aural and visual) as the autopilot disconnect. If the investigators heard the AP disconnect sound on the CVR this may also prove that the crew could not successfully engage the automatics, not that they disconnected it.

[de facto]

Yes FDR will know if stab was trimmed by AP or pilot trim switches.

[Sciolistes]

What if it is a trim runway and neither A/P nor manual trim switches were used. Is it a third state or an event without a corresponding trigger which has to be specifically interpreted as such by the individual(s) analysing the data?

[A4]

In response to FAB777

Airbus have designed their FBW system to be simple in the event of a go-around. For those not familiar, you simply firewall the thrust levers - AND IT IS IMPERATIVE THAT YOU DO FIREWALL THEM. This ensure that SRS (Speed Ref System) engages with appropriate FD pitch up guidance - newer FMGC standard will also engage NAV to follow the Missed Approach procedure (assuming FLT PLAN is sequenced....!) with AP engaged it is a non-event. In the event that FAB mentioned the aircraft simply did as it was being asked to do i.e. capture and descend on the glide....as that was the mode still active. Know your aircraft!

So, it amazes me that people still fail to push the levers till they won't go any further - how much simpler could you make the system? :ugh:

I'm not overly familiar with the Boeing, but from reading this thread it does seem that under certain circumstances the pilots workload can be (significantly) increased on a go-around depending upon whether it's a single/dual channel approach with pushing to overcome trim etc......

Good thread though.

A4

[Bus429]

The foregoing is one of the reasons I don't visit PPRune much any more. Those with absolutely no idea are making ridiculously wild conjectures, countered by the few who know (who then obviously give up, realising the futility of arguing with idiots). For God's sake, leave it to the investigators who will supply probable cause (unless, of course, a conspiracy theory emerges...)

[Kulverstukas]

Just published: MAK doesn't find any evidence of malfunctions after full decipher and analyzing of FDR and CVR recordings.

MAK-IAC investigations 737-800 A6-FDN 19.03.2016 (http://mak-iac.org/en/rassledovaniya/boeing-737-800-a6-fdn-19-03-2016)

The Interstate Aviation Committee (IAC) informs that the investigation team together with UAE and USA experts continues to work on the collection and analyzing of information. The accident aircraft fragments identification and layout works are going on at Rostov Civil Aviation Factory N412 facilities.

On base of the onboard recorders data analysis there were no failures of the aircraft systems or components, and power plant failures detected. The aircraft had valid Airworthiness Certificate, received all necessary maintenance and was in working order before the last departure.

The detailed analysis of all factors related to the aircraft operation is going on. The investigation team plans to conduct necessary examinations of the aircraft control system recovered components and elements. The Boeing Company received the request for providing of the technical documentation necessary for the evaluation of the aircraft systems performance, and also the information requested related to the Boeing aircraft events with similar scenarios.

At IAC Laboratory the works on the data preparation for the mathematical model analysis and for the aircraft flight trajectory recovering supported by the audio and video reconstruction of the accident flight.

In the context of the media information with some links to the Cockpit Voice Recorder (CVR) data, the Interstate Aviation Committee states that in accordance with International and Russian air accident investigation policy the CVR data is considering as the non-public disclosure information. The IAC has not disclosed and have no intention for the disclosure of this data. The work on the transcript of the flight crew conversation is going on. Today we have more than 1 hour of flight crew conversations transcribed, including the final stage conversations. Now in the IAC Laboratory experts from UAE, RF, USA, and France are working together on this task.



http://mak-iac.org/upload/iblock/bcb/IMG_1648.JPG

[BeGuidedAccordingly]

Translation of the above article frin 29 March:

The Interstate Aviation Committee informs that the commission of inquiry into the accident involving the United Arab Emirates and the United States experts continues to collect and analyze information. On the basis of the Rostov Civil Aviation Plant number 412, work is continuing to identify and "computation" fragments of the aircraft.

As a result of a preliminary analysis of flight recorders so far not revealed any failure of systems and components of aircraft, powerplants. The aircraft had a valid airworthiness certificate, passed the necessary maintenance at the time of departure was not defective.

Detailed analysis of information, aimed at identifying all the factors of the technical operation of the aircraft continues. The Commission plans to carry out the necessary investigations surviving units and elements of the aircraft control system. The Boeing Company has been requested to provide technical documentation necessary for the assessment of performance of the aircraft systems, as well as information on all the events with Boeing aircraft with similar scenario situations.

In the laboratory, the IAC continued work on preparing data for mathematical modeling and restoration of aircraft flight path with the audio-visual reconstruction of the accident.

In connection with the information appearing in the media containing data references a cockpit voice recorder, the Interstate Aviation Committee said that, in accordance with Russian and international rules of accident investigation, the information cockpit voice recorder (CVR) can not be disclosed. MAK not betrayed, and has no plans to bring this information to the public. Work on the "discharge crew talks" continues. By now deciphered more than 1 hour of negotiations, including the final stage of the flight portion. This work in the laboratory of IAC attracted UAE experts, the Russian Federation, USA, France.

[lapp]

The foregoing is one of the reasons I don't visit PPRune much any more.

So why have visited so far? What you were expecting? If you are only interested in the official report results, just check on the relevant websites every 3 months or so.

[Bus429]

I will, lapp, many thanks.

[flyingchanges]

PF pushing on yoke, PNF pulling on yoke, PF trims nose down to counter PNF forces on yoke. Fatigued, late at night, IMC, disoriented, this would be easy to do.

[Sciolistes]

Not so much as easy as unimaginable. The leak is an incomprehensible interpretation of events and people are swallowing it so readily!

[silvertate]

Furthermore, the 737Max has the same flight control system, the sole difference being the mechanical aileron - spoiler mixer unit being replaced by a flybywire unit that replicates the mechanical one with less weight. So, the system does meet modern certification.

The 737 is running on grandfather certification rights, and systems that remain the same do not have to be recertified. This does not mean these systems meet modern certification. Please correct me if I am wrong.

As to flying on stab trim. Sure. But most other aircraft have an intermediate option, between both elevators working and having to use the trimmer. It is called a proper control split. There is normally a switch or lever, and a simple pull separates one side from the other. So if a rhs control column, cabling, or elevator jams or freezes solid, the lhs system can fly the aircraft in almost the normal mode. Its called redundancy. And only after both elevators jam, do you need to use a trimmer. (And use the trimmer in the reverse sense, if it is a trailing tab trimmer. Which is confusing, but has been done for real in Manx.)

But in the 737 if one elevator jams or ices up, you lose the whole elevator system. That is not redundancy, that is 1950s technology still running 65 years later. It is like going to the car show room, and being shown a brand new Ford Popular with a new engine and dashboard. Would you buy it?


Ford releases their new 2016 model, following Boeing philosophy.
The new model has a new engine and uprated dashboard.

https://upload.wikimedia.org/wikipedia/commons/thumb/f/fd/RRT339_020510_CPS_%284576174140%29.jpg/640px-RRT339_020510_CPS_%284576174140%29.jpg

[donotdespisethesnake]

Question from the peanut gallery:

In an FBW system the FDR can record captain/FO inputs separately as a matter of course.

For 737 type, what inputs would normally be recorded on the FDR with respect to control columns, and what is recorded in the event of physical separation of some or all control column functions?

I am wondering if MAK could have hard evidence of pilot disagreement from FDR, or whether they would have to deduce that from the voice transcript.

[Kulverstukas]

I am wondering if MAK could have hard evidence of pilot disagreement

I doubt if MAK yet told anything about such things as "pilot disagreement", "quarrel in the cockpit" or "chilling nonhuman cries" at all.

[donotdespisethesnake]

I doubt if MAK yet told anything about such things as "pilot disagreement", "quarrel in the cockpit" or "chilling nonhuman cries" at all.

Understood. I take the leaks with a big pinch of salt.

I am wondering what is theoretically possible to determine conclusively whenever the final report is made. Some incidents of dual inputs have been made on "the balance of probability", and not all parties agreed to the conclusions. In the case of AF447, the FDR showed in black and white there were dual (and opposing) inputs, which is possible to see because of FBW controls.

It seems to me that in 737 the control surface deflection can be recorded, but not individual control column movement which would definitively show conflict of elevator command.

[DiamondRider]

Understood. I take the leaks with a big pinch of salt.

I am wondering what is theoretically possible to determine conclusively whenever the final report is made. Some incidents of dual inputs have been made on "the balance of probability", and not all parties agreed to the conclusions. In the case of AF447, the FDR showed in black and white there were dual (and opposing) inputs, which is possible to see because of FBW controls.

It seems to me that in 737 the control surface deflection can be recorded, but not individual control column movement which would definitively show conflict of elevator command.

The FDAU records parameters for the CAPT and FO control columns. Both force (push/pull) and column deflection, so conflicting inputs will easily be discovered. Parameters for the individual elevators and the aft input torque are also recorded. In addition to this there are also other sensors recorded that could be useful, Neutral Shift Sensor, Mach Trim Actuator and several stabilizer sensors. Investigators should be able to tell what was going on with the pitch control.

[mickjoebill]

I am wondering if MAK could have hard evidence of pilot disagreement from FDR, or whether they would have to deduce that from the voice transcript.

In cockpit camera, a human performance recorder!

Mickjoebill

[Bus429]

Much as I hate to add to all this, I am looking at a DFDR Service Report (readout) for a 737 CL that shows control column position, control wheel position, rudder pedal position and the positions of the control surfaces connected to these devices. No reason to think an NG wouldn't be recording the same.
(Please note, I am not speculating as to the cause of this tragedy, just the comments about technical aspects of the 737).

[LandIT]

Russia’s Interstate Aviation Committee (IAC) said a preliminary analysis of the flydubai Boeing 737 data recorder has found “no onboard system malfunctions, aircraft components’ defects or power system failures.”

Seems to indicate some mishandling (at least) no doubt with extenuating circumstances in that weather at that time of night. Let's wait for the real facts to come from the investigation. Speculation is no longer useful.

[ACMS]

He's the boss of the Aviation regulator and the Chairman of an Airline in the UAE???? You are kidding me.......:mad:

Well at least they are obvious in their intentions!!

[neville_nobody]

The 737 is running on grandfather certification rights, and systems that remain the same do not have to be recertified. This does not mean these systems meet modern certification. Please correct me if I am wrong.

You are correct the 737 type certificate is dated in 1967 with 55 revisions! You can look it up on the FAA website.

[rainbow gravity]

such concrete information that one pilot was pushing CC and second was pulling back IMO couldn't' came out of nowhere. I mean reporters have tendencies to overdraw, boost, exaggerate, but this - if actually not true - would be a pure facts inventing, a simple lie.

According to avherald today:
« On Mar 29th 2016 the MAK condemned all “leaked” information reported by Russian media as false and stated, that no information whatsoever has been leaked from the investigation. About 1 hour of CVR has been transcribed so far, including the crew communication in the final stages of the flight. Mechanical reconstruction as well as preliminary analysis of the flight data recorder so far do not suggest any malfunction of the engines of the aircraft and no failures of aircraft systems or components, the aircraft had all necessary documentation and certificates of airworthiness, passed all required maintenance and was airworthy at the time of departure. The MAK is now undertaking mathematical modelling of the aircraft's flight trajectory in combination with the sounds available on the CVR. The identification of wreckage pieces and mechanical reconstruction of the aircraft continues. »

[DiamondRider]

Well, there are force sensors on each column. Between the column itself and the Control cables. If one pilot pushes on the column and the other pushes you will see this as positive and negative values. No need to activate the breakout mechanism to see conflicting inputs.


We are now talking about the force exerted on each column, as you say the column deflection will be the same (who is strongest) until the breakout mechanism is activated.


And yes, the sensors are in the AMM.

[FDMII]

Depending upon the data frame, (1, 2, i, 4-4A, 5, 6), for the CL they may have just the left control column and wheel position, or both left & right control column and wheel position and I believe just one data frame also has control column & control wheel force.

As may be deduced from discussion above, this will obviously be different (more elaborate) for the NG due aircraft manufacturing date and FAA data frame requirements (http://www.faa.gov/documentLibrary/media/Advisory_Circular/AC%2020-141B.pdf).

If the recorder for the NG is the same as the CL and if it is FDAU-equipped but with no DMU then the QAR will collect the same data that is going to the SSFDR; if FDAU and DMU-equipped, a separate, more complex data frame with higher resolution and sample rates, (still proprietary and very expensive), is possible but as some here will know, the QAR is not crash-protected.

[Infieldg]

As a computer programmer who found this forum invaluable when MH370 went missing but from MH17 to Metro has seen it get more and more diluted with accountants who think they can fly better on their PC, could you perhaps make 'profession' and 'pilot's licence Y/N' mandatory and display them in the left sidebar? Because any fool can google, and it'd be nice to see 'meteorologist' next to a post on windshear instead of 'accountant'.

Banning non-pilots would lose the valuable input that controllers and ground crew and engineers and investigators have to offer. But the flight sim guys are really killing you :(

[FDMII]

Infieldg, those who know their stuff and actually do this work (and who remain to post here despite the high noise levels), recognize good work and engage it and give the rest a miss. Always, there are at least two levels of dialogue going on at once! ;-) Having posted here for over a dozen years now, I don't think it's going to change and as you say, you can't just limit it to professional pilots - too much would be lost!

[framer]

@Infieldg, it had already been diluted massively by the time you joined. You would have loved it ten years ago. I have learnt a lot over the years from pprune. Now, I worry that pilots ten years behind me will believe half the rubbish that is posted here.

[klintE]

The FDAU records parameters for the CAPT and FO control columnsTo be strict, FDAU doesn't record anything, it collects signals and transfer them to recorders or other devices.
But you're right, DFDR from 5 yo Boeing a/c, must be able to record CC inputs separately.
Screen below comes from report related to TK1951 (737NG) crash during landing at AMS

http://pic.al/HXc.png

http://pic.al/FXc

[Sciolistes]

To be strict, FDAU doesn't record anything, it collects signals and transfer them to recorders or other devices.
But you're right, DFDR from 5 yo Boeing a/c, must be able to record CC inputs separately.
Screen below comes from report related to TK1951 (737NG) crash during landing at AMS

http://pic.al/HXc.png

http://pic.al/FXc
How did the Capt and FO separate the controls to such a degree without breaking out?

[m39462]

Sciolistes, there are 3 curves in that plot, not 2.
The two for the control column positions are coincident, the third that differs matches the scale in the right margin and gives pitch displayed in the captain's ADI.

[FDMII]

Yes, correct, thanks, KlintE.

[Sciolistes]

Sciolistes, there are 3 curves in that plot, not 2.
The two for the control column positions are coincident, the third that differs matches the scale in the right margin and gives pitch displayed in the captain's ADI.
Ah, that's my earlier point. The Russian leak said the pilot were applying opposite elevator commands. My point is that simply cannot be seen on the FDR without breaking out.

[archae86]

My point is that simply cannot be seen on the FDR without breaking out.
According to CFR 135.152 (https://www.law.cornell.edu/cfr/text/14/135.152) Flight Data Recorder requirements, force recording is definitely contemplated.

In particular this reference includes (88) All cockpit flight control input forces (control wheel, control column, rudder pedal).

In my experience a typical accident report only displays a very small subset of the recorded parameters.

Force input reporting is not only possible, but in at least some portion of the fleet, is required.

[FlightDetent]

Force applied to the control column is recorded too.

[FGD135]

But I am referring to the leak which said that the captain and of where applying opposite forces.There has NOT been any leak which has said this. This idea of "dual inputs" has come purely from media speculation. That media speculation would be entirely due to these words, leaked from the CVR:

Don't do this ... don't do this ... don't do this ...

Most of us here realise that these words were most likely directed at the aircraft, rather than the other pilot, but that did not stop the media speculation, which began with a simple "pilots disagreeing" turning into "pilots fighting".

We must understand that the media are under pressure to keep up the daily reporting, despite the absence of any new material.

For us here to spend so much time on this "dual inputs" speculation not only makes us as bad as them, it also encourages them.

Can we please drop the "dual inputs" stuff and stop encouraging the media flights of fancy?

[Pilot DAR]

May we have discussion about facts, and not speculation about "leaks"....