B737 @ Aberdeen

[olster]

There has been a well documented report in the popular press of a botched B737 go around @ABZ relatively recently. Subject to an initial AAIB report and curious radio silence around here. The modern 737s have a history of go arounds that have had less than optimum outcomes. The problem lies in the integration of the automation into an old technology airframe. If you know what you are doing and have good knowledge of the autopilot / auto throttle interface coupled with config / position on approach considerations then most go around scenarios can be handled proficiently with less ‘startle’ etc. Would be interested to know what happened for the purpose of education.

[Airbanda]

I too was surprised not to fond comment here sooner. The AAIB Special Bulletin stating the facts as known is here:

https://assets.publishing.service.gov.uk/media/61a73d1e8fa8f5038358c0d8/S2-2021_Boeing_737-8K5_G-FDZF.pdf

As a non pilot but long term observer of such things it does look a bit dicey given the altitude and rate of descent. Professional commentary will be interesting.

[meleagertoo]

Penny to a pound it'll come down to "Read your FMAs!" . Do that and it's hard to get it wrong.
18 seconds from the instruction to go around and actually implementing it! Even though forewarned? Wow!

[ATC Watcher]

To be fair and correct he was not instructed to "go around" but to "break the approach" at 2600 ft , or 2400ft above runway threshold, an quite some NM out, so not a very urgent maneuver. But the problem does not lies there.

[Jwscud]

If you don’t need to go around immediately (ie at DA/close to the ground) taking a few seconds to confirm everything’s set correctly and how you’re going to do it with your colleague is good airmanship, particularly if you are very close to them missed approach altitude.

[FlightDetent]

Above 2,000 ft radio altitude, one press of a TO/GA switch commands thrust to the full go-around N1 limit (although this is not included in the Flight Crew Operating Manual and was unexpected by the crew)
Funny that.

[172_driver]

Penny to a pound it'll come down to "Read your FMAs!" . Do that and it's hard to get it wrong.
18 seconds from the instruction to go around and actually implementing it! Even though forewarned? Wow!

That's the standard quote from the training department. I respectfully disagree. Not saying the FMA:s isn't import, but much better is knowledge about pitch/power and ability to execute such in a high workload environment. Getting yourself into a dive at 3000 fpm and 280 kts means 1. You are overloaded, 2. Interpreting FMAs isn't going to happen.

The 737 is a baby to go-around in if you're one step ahead, like the one-push reduced G/A thrust. It's a nightmare if you're one step behind as you have to run the trim like crazy if you're turning, levelling off, reducing thrust simulatenously.

I feel sorry for the crew having to read about themselves in the newspaper.

[Banana Joe]

There it is, 60's technology for you. I am not saying pitch and power isn't important, but the ability to keep the autopilot engaged at the beginning of the Go Around would make dealing with the workload less of an hassle.

[UV]

The FO had been off flying for 11 months and had two flights with a Trainer in the previous week. This was his fourth flight in those 11 months, so I assume his second after “release”.
Unfortunately this report does not say who was the handling pilot.

[His dudeness]

Being a lowlife businessjet pilot, I always wondered about our training:

ALWAYS full power etcetc. At 2000ft plus there is no need for that. Apply some power gently, bring her nose up gently, raise whatever has to be raised and go about your business....

[zero/zero]

meleagertoo

Another one of those incidents that’s all too easy to criticise from the comfort of your armchair at home at 1g.

Truth is go-arounds are insufficiently practiced (I’ve done 1 real one in 5 years) and when they are, it’s generally in the sim from DA and you know it’s coming. Both Airbus and Boeing have some pitfalls with Intermediate GAs (close to MAA or above it) and the 737 adds the complexity of disconnecting the AP on TOGA selection and giving you a load of pitch/power etc etc.

As with all accidents, it can never be distilled into one cause. It’s been a crazy 1.5 years in aviation and many of us feel rusty. In this case they had a bad day in the office, but recognised and recovered the situation.

[PAXboy]

As PAX I want to ask, wth some SIM sessions, do they not just get you 'depart and get to cruise' to nominated destination. THEN they throw things at you? Or do you always know what is about to happen?

[zero/zero]

Sim time is very precious/limited (4hrs x 4 times/year) so there just isn’t time for a “depart and cruise”, which would honestly be a pretty wasteful use of the time. It’s generally a very hectic box ticking exercise to get all the mandatory manoeuvres ticked off - engine failure, go-around, 3d approach, 2d approach, low viz etc etc. and then training for stuff that comes around periodically - upset recovery, GPWS, depressurisation… and the list goes on.

There is an element once a year called a Line Operational Evaluation which is a small company scenario with some failures thrown in. Maybe 1hr of one of the 4hr blocks

[wiggy]

PAXboy

Answering your first question first…generally these days no, you don’t know exactly, but sometimes, depending on what chunk of recurrent checking training you are doing you can make a guess…for example if you have just started the Low Visibility training part of a session you know there’s a high probability there’s a rejected take-off coming up….

Other than that as has been said given the requirement for mandatory items to be got through sim time is tight/precious…

As a result you’re probably you’ll never get a standard start up, departure and be allowed to get all the way up to 30,000 feet plus before the wings start falling off :}

You’ll either get ( thinking Long Haul sims ) a scenario where the failures are chucked in on a short sector, e.g. Bahrain/Doha or Heathrow to Manchester…that would be the Line Orientated Evaluation that zero/zero mentions…..or you (virtually) go off from London heading for the Middle East but the problems start immediately the gear is up, if not before…and you end up back in London having never got East of Dover….

If an exercise needs to be commenced at cruising level (e.g. depressurization and rapid descent..or maybe a problem that starts in the cruise such as a fuel leak) you’ll be fast forwarded to start that chunk of the lesson up there, so to speak…

[ACMS]

This was the ALT-GO procedure we used and it works very well. Slows things down nicely and allows time to think.

https://cimg5.ibsrv.net/gimg/pprune.org-vbulletin/1058x1004/6d4929df_3601_4b6e_94e3_532c0ae8f9fe_c5d3fd0845e67e3b72da988 1ec88dbd1a8ec9075.jpeg

[Denti]

Yes, that is the nice difference between airbus and boeing. On the 737 there are only limited methods to get out of the approach mode: change the ILS frequency, TOGA, or disconnect the autopilot and reset the FDs. That said, in the airlines where i flew the 737 every ILS was flown dual channel, which allows for automatic go around without autopilot disconnect, which makes the whole thing a lot more relaxed.

Had an ATC advised go around this summer, since there was no urgency i took the time to do a short brief what i will do next, and then calmly did it. Quite often there is no real urgency during an intermediate go-around and a few seconds waiting helps a lot. During my career i was lucky enough to had to do roughly one go around per year on the line, and most of those were intermediate ones, not the normal training case of one engine inoperative at minimum or below minimum (balked landing training).

[FlightDetent]

PAXboy

Licence revalidation checkride
- single-engine ILS
- single-engine VOR
- single-engine G/A
- EFATO w/ max x-wind
- max x/wind dual eng apch without FDs
- dual hydraulic failure
- emerg elec (backup gen only)
- emergency DES
- probably FAC1+2 with single engine
- ground evac due fire
- TCAS excercise
- WSHR recovery
- GPWS escape
Those are the pretty much the mandatory exercises. Apart from the Emerg. DES and WSHR all manual flying.

[Uplinker]

Firstly my sympathy to the crew. They had probably been up at "oh-god o'clock", and doing a Palma W pattern can be a frazzling experience, especially after almost a year of not flying.

I think there are two aspects to this. First; I know that both pilots had been in the SIM, but a concert violinist would never be expected to perform a difficult musical passage well unless they had had days or even weeks to practise it at home, over and over beforehand. Unfortunately, we are not allowed to take the aircraft home to practise!
Even in the SIM, one never has the luxury of practising go-arounds say five times in a row in order to hone our responses. Usually even if it was wobbly and untidy; if the manoeuvre is flown just about within the limits then the box is ticked and the session moves on.

Secondly the 737.......There will be those shouting at the screen: "just fly the got-damned plane! Are you a pilot or an idiot?". However the limited 1950's technology of the B737 and the pitch-power couple from the engines demands a hefty push forward on the yoke and a trim forward during the go-around transition, which is opposite and counter-intuitive to what one would expect when wanting to go up. If you fly a go-around every day or every week; fine, but once a year ?

Why Boeing have still not introduced FBW or decent autopilots and auto-thrust that can handle all flight phases to the 73, baffles me. They could even have brought in FBW just in pitch, which would have made a huge difference enabling auto-trim, and things like bigger engines and MCAS would have been very easy and safe to implement.

[Banana Joe]

Denti

This would work, but with one caveat: the AP must be disengaged by 350 ft RA to avoid having an aircraft severely out of trim.

[Denti]

Well, true in one way. Although even in the initial type rating and line training it was specifically flown to 200ft on two, click it off and land it. Yes, one has to be aware of the (unfortunate) trim, but it is very much manageable. In normal line life it was usually disengaged well before that, except for autolands of course, which both pilots could do whenever they wished to, on CAT III capable runways.

That said, the main culprit in many ways is, outside of EBT, the fact that there is the usual litany of standard stuff one has to do every simulator which is basically box ticking but does not really help all that much on the line. Like for example intermediate all engine go arounds, which are rarely if ever practiced.

[Alrosa]

Hopefully we can all learn lessons from the incident, once the full report is published. Probably worth including relevant actions in the approach brief when operating into airfields with particularly low MAAs…ESSA comes to mind, but I’m sure there are many others (perhaps not directly relevant to this incident, but an example of a situation which requires a little thought).

The sad reality is that anything out of the ordinary has a tendency to catch us out, regardless of experience.

[Brian Pern]

Banana Joe

Unless you are in IMC, I fail to remember a time when I have left the autopilot in that low, I always disconnect roughly 1000agl, regardless of the wind/wx, it helps you get a 'feel' of the aircraft, I know some companies don't like you disconnecting too early but 350ft......

This type of go around is always going to be an issue, light aircraft, low level off etc, although its easy to criticise the crew of this, we all could have been there.

In recent years, I have had the pleasure of flying many times with a low landing weight 737, I find during the brief discussing the G/A Power setting helpful. In this case where I would break off an approach, I would find the old philosophy of Automatics in or Automatics out more helpful, I would be much happier setting the power manually. but then I am a dinosaur.

[parkfell]

Easy to sit back in one’s armchair & pontificate.
An unusual G/A which was probably never encountered before on the 737. Probably just read about it in the dim and distant.

One thing to bear in mind is that once TOGA is pressed the single channel AP disconnects and you are now handing flying, with increasing thrust and pitching up which needs to be controlled with application of forward trim.
Flap 15 retraction to say F5/F1/UP with pitch down resulting. This is particularly noticeable from F1 to flap UP. Significant back pressure is necessary with trim in a timely manner. Lack of scan with low situational awareness.
Unless this is regularly practised, it doesn’t come as any great surprise that it catches inexperience pilots out. No mention who was PF. I would speculate the FO…?? Add to this the lack of flying. Was the PM monitoring effectively….? Also caught out…
Emirates SOP (was?) leave the flaps at F15 until platform altitude. Then retract having ‘bugged up’.
Reduce the workload by timely reengagement of the AP once trimmed.
All easier said than done in the comfort of the armchair.

What will be interesting is the response of the Training Dept…

[Rt Hon Jim Hacker MP]

The problem for the training department is satisfying the requirements of the LPC. A "test" that is so out-dated it is not fit for purpose. It hasn't been relevant for about 30 years or more. I went to a training conference just after the 787 went into service. A photo of a 787 flight deck with HUD was presented next to that of a DC-3. The one thing they had in common was the check schedule to revalidate the type rating. Until the regulators get their heads out of their arses mishaps like this will continue to happen. Anyone out there ever actually had to hand fly a single engine ILS and G/A for real? Yet we will all do it hundreds of times in a career in the sim. A total waste of time.

[meleagertoo]

Well, some years ago my company (a particularly garish one) realised, possibly following a similar incident, that the most commonly cocked-up manoeuvre was the two-engined medium level go-around. hardly surprising as it was only ever practices once on type conversion and afair never seen again - and here is that very same manoeuvre cocked up...
I seem to recall that Mr Boeing built some pitfalls into this seemingly simple execise though my menory is clouded with subsequent Airbus time and which procedure was which.

My 737 training emphasised rigidly that a/t was NEVER to be used a/p out - either both in or both out. You either flew manually or automatically, one or the other - and the other one was to learn to fly 'through' the f/d if you got in a pickle. Fluffing your mode selections and/or failing to read FMAs could be very distracting indeed and it was necessary to learn to fly manually while ignoring the flight director.
I wonder how many are taught that today?

But then, I was taught the see-saw runaway trim detail that could just (perhaps)have saved the Max crews too, had they been aware, or rather awake enough by use of basic airmanship never to have gotten into the problem so deep that they neeed it..

[FlyingStone]

Why Boeing have still not introduced FBW or decent autopilots and auto-thrust that can handle all flight phases to the 73, baffles me. They could even have brought in FBW just in pitch, which would have made a huge difference enabling auto-trim, and things like bigger engines and MCAS would have been very easy and safe to implement.

Two words: type commonality.

To be honest, yes, 737 has lots of issues, but on the other hand, a go-around at 2000ft AAL on ATC request should hopefully be a non-event for a professional crew in any airplane type. There's plenty of time to discuss the actions, verify missed approach altitude and go about it nicely, slowly and methodically.

It's the pushing of TOGA for dear life when there's no immediate threat that gets people in trouble.

[blue up]

Airline I worked for until a couple of years ago had an enterprising individual who would sell you a copy of the TRE notes for the Sim session. You would always know exactly what was going to happen.

[Alrosa]

FlyingStone

Whilst I don’t disagree with the theory, recent events (this incident, the French Bee A350 at ORY - and others) would seem to suggest that non-standard GAs are not “non-events” for experienced crew. So I think one of the questions we need to ask is : WHY ?

[zero/zero]

FlyingStone

No easy solutions... many training departments try to take away the grey areas by simplifying the procedures and having one GA procedure regardless of where you initiate it. As we've seen; this is particularly pertinent on the Boeing where it's not necessarily easy to get out of approach mode

[MissChief]

Why single-channel AP on the ILS?

[parkfell]

Why single-channel AP on the ILS?

Flying a CAT 1 ILS with conditions above minima for CAT1…

As previously stated, a dual channel AP approach (CAT3) results in autotrimming up at around 350’RA.
You could fly Dual CH, remembering to discount prior to this trimming, or dealing with an out of trim ac below this point once the AP is disconnected.

[Uplinker]

QUOTE=FlyingStone;11151846]Two words: type commonality..[/QUOTE]

Fifteen words: Trying to maximise profits rather than improve their aeroplane and bring it up to date. :)

I agree with your other point though, I cannot remember the Boeing 737 FCOM now, but surely a discontinued approach from well above DA need not employ TOGA; just a level off then a gentle climb and clean up.

[lederhosen]

Yes you can certainly use vertical speed above a height which I no longer remember as I have used it in earlier days on a 737 discontinued approach. I expect Denti will know.

[FlyingStone]

Dual channel autopilot go-around isn't available on the 737 unless both autopilots are engaged and FLARE armed is announciated, which won't be until after the self-test is completed below 1500ft RA.

[midnight cruiser]

Question for the 737 pundits - can avoiding using the toga button for aborting an approach, cause complications for the FMC logic/legs/route sequencing?

[Banana Joe]

No, one of the conditions to sequence to the missed approach segment is a RoC above a certain value specified in the FCOM.
​​​​​​Just disconnect the AP, ask the PM to recycle flight directors and engage LNAV or HDG SEL and LVL CH. Around 75% N1 with 7,5° nose up attitude should establish the aircraft in a gentle and manageable climb with flaps 15 and gear up and gives the PM time to rebuild automation.

[midnight cruiser]

Thanks. I guess that's what it was - we stayed in VS to gently capture the MAA, either from above or below, and I momentarily had problems joining the route up for another approach in the FMC (I forget how we got around it, or maybe it was fat finger syndrome; it just had me scratching my head for a moment).

[Jump Complete]

No, one of the conditions to sequence to the missed approach segment is a RoC above a certain value specified in the FCOM.
​​​​​​Just disconnect the AP, ask the PM to recycle flight directors and engage LNAV or HDG SEL and LVL CH. Around 75% N1 with 7,5° nose up attitude should establish the aircraft in a gentle and manageable climb with flaps 15 and gear up and gives the PM the time to rebuild automation.

Thanks for that. Fairly new to the 737 (about 300 hours over 22 months thanks to Covid and furlough) but that’s what my thought process was. TOGA making things far too exciting in that scenario. One potential gotcha that occurs to me with the relaxed go-around though, is missing the normal sequence, eg
‘Go-Around. …Flap.(if appropriate)‘
’Positive Climb’
’Gear Up’
Similar to a wind-sheer escape where we brief that once clear (obviously without changing configuration) we will then make the ‘Go-Around’ ‘Flap 15’ call to get into that mindset and ensure we don’t forget to raise the gear etc.

[Brian Pern]

Why take the autopilot and A/T out?

De tune ILS freq, (A/P will revert to CWS P and R)
Press HDG SEL
V/S +700 ft per minute
Don't forget a appropriate speed in the IAS window
The automatics should take care of everything else

Its the old KISS priciple

[Banana Joe]

Why take the autopilot and A/T out?

De tune ILS freq, (A/P will revert to CWS P and R)
Press HDG SEL
V/S +700 ft per minute
Don't forget a appropriate speed in the IAS window
The automatics should take care of everything else

Its the old KISS priciple
Did it once on a Classic and the behaviour was not as advertised, the autopilot disconnected. I may give it a try again the next time.

[Brian Pern]

Did it once on a Classic and the behaviour was not as advertised, the autopilot disconnected. I may give it a try again the next time.

The joy of the Classic, well it can do strange things.
It should drop into CWS, but of course if the A/P drops out, just get the PM to reprogram the MCP.

May be worth asking to to try it in the SIM when you get time, Next time I have time, i'll try it as well!!

The most important thing is not to panic of course.

[max alt]

There is no doubt that a lack of currency played a major part in this incident,the fact that the aircraft was essentially out of control close to the ground makes us ask ,how this could happen?.I have never been an advocate of the move from 6 month checks to yearly ones as any cost benefit is outweighed by a dread/fear of passing your check .The simulator should be a booster to REINFORCING your competency .A standard two engine missed approach should now be regulated into recurrent training.The language could be changed to refer to "break off your approach,climb 3000,turn left heading etc either demonstrated or brought in by SOP that below a certain altitude say 1000,TOGA is applied ,whilst above other modes can and may be used.I remember this being practised as a result of a similar incident some years ago and I am saddened to see the message has been lost.Just a personal view from an old trainer.

[BizJetJock]

No need to change the language. From the AAIB bulletin:
At 2,600 ft amsl the aircraft was instructed by the radar controller to break
off the approach, climb to 3,000 ft and turn left onto a heading of 270°

[Uplinker]

max alt said: .."standard two engine missed approach should now be regulated into recurrent training."

Yes, especially on non-FBW planes with underslung engines; where there is a large manual trim change requirement during the transition.

[BizJetJock]

max alt said: .."standard two engine missed approach should now be regulated into recurrent training."
It already is and has been for quite a few years now. If your operator is not doing it then they are not compliant with AMC1 ORO.FC.230
https://cimg6.ibsrv.net/gimg/pprune.org-vbulletin/618x249/screenshot_2021_12_08_110357_936fcf3eedd8f593899ede2ccc46844 078e977de.png

[cessnapete]

[QUOTE=Denti;11151611]Well, true in one way. Although even in the initial type rating and line training it was specifically flown to 200ft on two, click it off and land it. Yes, one has to be aware of the (unfortunate) trim, but it is very much manageable. In normal line life it was usually disengaged well before that, except for autolands of course, which both pilots could do whenever they wished to, on CAT III capable runways.

That said, the main culprit in many ways is, outside of EBT, the fact that there is the usual litany of standard stuff one has to do every simulator which is basically box ticking but does not really help all that much on the line. Like for example intermediate all engine go arounds, which are rarely if ever practiced.[/QUOTE



Am I being simplistic in reaction to this incident? Seems a lot of heads down and button pressing.
Was the HP incapable of disconnecting the automatics and smoothly applying enough thrust, ( ie push the thrust leavers forward and raise the nose) while making a gentle climb and turn to the requested heading and altitude. The NHP meantime resets the automatics to reengage the autopilot when required.

[172_driver]

Am I being simplistic in reaction to this incident? Seems a lot of heads down and button pressing.
Was the HP incapable of disconnecting the automatics and smoothly applying enough thrust, ( ie push the thrust leavers forward and raise the nose) while making a gentle climb and turn to the requested heading and altitude. The NHP meantime resets the automatics to reengage the autopilot when required.

There's almost no point stating the bleedin' obvious anymore. The only talk on this thread is about which way manipulating the automatics is easiest to get the job done. It's a patch work where something essential is missing! Adding a bit of thrust and raising the nose a couple of degrees is such basic instrument flying I am sure an early IR student would sort it out better... as he wouldn't even know how the automatics work.

[Denti]

Of course the issue could have been saved both by competent manual flight as well as competent use of automatics albeit with some more work as it is after all a boeing. That said, in my experience in various airlines over the years, especially british based airlines do have quite a heavy focus on automation at all times, which i have not seen nearly as much in continental ones. Especially now after the pandemic to battle the perceived and real loss of flying skills, which in fact does not help there at all as it simply reinforces the loss of manual flying abilities.

That said, i had a similar situation just yesterday, go around from a fairly high altitude on approach into TLV (ordered by ATC due to a drone in the approach path), but in an A320. A complete non-event, push thrust levers to TOGA and pull them back into CLB, while the PM retracts the flaps by one step, no autopilot disconnect (but again, of course both autopilots were in, same as when i was flying the 737). In this case the aircraft design helped, but so did having done the same earlier in the year both in the simulator and on the line. So we climbed back the few hundred feet and then got vectored for an approach on another runway.

[Uplinker]

I am not going to get into the "why fly on automatics, aren't you a real pilot" debate - there are plenty of valid reasons both for flying manually or with the automatics - but having to take manual control if something vaguely tricky comes up because the auto-pilot and the auto-thrust can't handle it is a poor option, and might explain why these things sometimes go wrong.

Please, Boeing, update your aircraft ! (And not with things like MCAS; it needs something decent).

[SOPS]

We all know the 737 needs a complete clean sheet redesign. I flown 737 2 3 4 7 and 8. Nice aircraft for its time ( the 200 series) .. since then it’s been ad ons to an old airframe. That being said, the 800 is a nice aircraft… but it could have been a excellent aircraft.

[Banana Joe]

My understanding is that wide bodies (and 757) are much better in terms of automatics. The only dinosaur in the room is the 737, including the MAX. Obviously I stand to be corrected on the 757/767.

[meleagertoo]

Am I alone in becoming most uncomfortable at the ever-increasing tendency to blame the aircraft or manufacturer when pilots fail to manage long-established systems they are supposed to know and understand? Many aircraft have systems or procedures that may not appear to some to be logical or intuitive but are we not given specific training on how to operate these? Is this not why we are paid as pilots?

737 pilots are supposed to know their modes, selections and expected outcomes just as much as Airbus ones are; if a discontinued approach is managed in a different manner between the two is one of them wrong, or a 'worse' design? Why does one manufactureer 'need a complete clean sheet redesign' when it works just fine - as long as you use the right drills? There are even sniffy remarks above about the pitch-thrust couple in aircraft with underslung engines as though it is something so undesireable it needs designing out. (With what? Another MCAS type masking device? Is that really progress?). Is this really something that bothers modern pilots? If so it a very scary attitude, being phased by the simple inherentand benign charactersitics of your particular type. There seems to be an extraordinary mindset about nowadays that if something works less intuitively on one fleet than another then there is something wrong with the design. There's one helluva lot about Airbus that is very, very far from intuitive, and I'm not talking about the mangled and sometines incomprehensible language in the manuals.

If it takes 18 seconds to respond to an instruction to break off an approach in any type I respectfully suggest it isn't likely to be the aircraft or designer's fault. Had the crew been familiar with the correct procedure the manoeuvre would have started in maybe five seconds. The timescale implies they didn't. That is nothing to do with the manufacturer or the systems. It is an operator matter. What of it if the autopilot reverts temporarily to another mode? It presents no hazard, that's just the way the system works - but it does presuppose the pilots understand a) what is happeneing and b) what they are doing about it.

Boeing Bashing took on fever pitch over the MAX accidents which once again were caused in the immediate sense by just this, pliots who appeared lacking in their systems knowlege. It is all very well (and correct) to blame Boeing for providing the root of the problem but the nub of the matter durng the event was that both aircraft were perfectly flyable had the correct drills, not to mention the most basic of airmanship been applied. I can already hear the shrieks of fury at such herecy, but for all that it remains absolutely true.
Nonetheless, had the ABZ event ended in a smoking hole the Media and this forum too would, I assure you, be ablaze with the most wild and hysterical accusations of Corporate butchery and gross irresponsibility yada yada yada on Boeing's part for so misdesigning the autopilot (once again). That would neither be fair nor rational, but you can be assured it would have happned. But does the difference that no one was hurt make the design of an autopilot that has passed the rest of time reputation unsullied for decades suddenly a lethal trap? No it dosn't. So while not suggesting we ignore any shortcomings let's also not condemn something that actually works perfectly well as long as the operator understands it. Just like every other aspect of aviation you can think of.

I see this inclination to blame the designer as an extension of the Children of the Magenta Mind philosophy or mindset where pilots and commentators seem increasingly unable to deal with matters that the aircraft doesn't resolve or interpret itself, ie a reliance on pavlovian response to specific conditions/warnings without much if any analytical process to back things up where there isn't a Big Red Light that says "XYZ Failure" with the corresponding book-reference requiring "ABC" as response. ECAM is the ultimate example of a system that encourages this mindset, and indeed all but prevents independent action - that damn nearly caused the total loss of a 380 didn't it? Had that crew been merely pavlovian button-pushers they wouldn't have survived.

What we've lost here is something called "Airmanship" (anyone still remember Airmanship?). Knowing and understanding the how and why of your systems, not just which button to press. Knowing your drills and procedures. Thinking through all actions.
Noy relying in the damned autopilot to get you out of every pickle and blaming the manufacturer when it doesn't. We are clearly increasingly becoming reactive operators as oposed to proactive ones.

It isnt, imho, a healthy trend.

[zero/zero]

Am I alone in becoming most uncomfortable at the ever-increasing tendency to blame the aircraft or manufacturer when pilots fail to manage long-established systems they are supposed to know and understand? Many aircraft have systems or procedures that may not appear to some to be logical or intuitive but are we not given specific training on how to operate these? Is this not why we are paid as pilots?

737 pilots are supposed to know their modes, selections and expected outcomes just as much as Airbus ones are; if a discontinued approach is managed in a different manner between the two is one of them wrong, or a 'worse' design? Why does one manufactureer 'need a complete clean sheet redesign' when it works just fine - as long as you use the right drills? There are even sniffy remarks above about the pitch-thrust couple in aircraft with underslung engines as though it is something so undesireable it needs designing out. (With what? Another MCAS type masking device? Is that really progress?). Is this really something that bothers modern pilots? If so it a very scary attitude, being phased by the simple inherentand benign charactersitics of your particular type. There seems to be an extraordinary mindset about nowadays that if something works less intuitively on one fleet than another then there is something wrong with the design. There's one helluva lot about Airbus that is very, very far from intuitive, and I'm not talking about the mangled and sometines incomprehensible language in the manuals.

If it takes 18 seconds to respond to an instruction to break off an approach in any type I respectfully suggest it isn't likely to be the aircraft or designer's fault. Had the crew been familiar with the correct procedure the manoeuvre would have started in maybe five seconds. The timescale implies they didn't. That is nothing to do with the manufacturer or the systems. It is an operator matter. What of it if the autopilot reverts temporarily to another mode? It presents no hazard, that's just the way the system works - but it does presuppose the pilots understand a) what is happeneing and b) what they are doing about it.

Boeing Bashing took on fever pitch over the MAX accidents which once again were caused in the immediate sense by just this, pliots who appeared lacking in their systems knowlege. It is all very well (and correct) to blame Boeing for providing the root of the problem but the nub of the matter durng the event was that both aircraft were perfectly flyable had the correct drills, not to mention the most basic of airmanship been applied. I can already hear the shrieks of fury at such herecy, but for all that it remains absolutely true.
Nonetheless, had the ABZ event ended in a smoking hole the Media and this forum too would, I assure you, be ablaze with the most wild and hysterical accusations of Corporate butchery and gross irresponsibility yada yada yada on Boeing's part for so misdesigning the autopilot (once again). That would neither be fair nor rational, but you can be assured it would have happned. But does the difference that no one was hurt make the design of an autopilot that has passed the rest of time reputation unsullied for decades suddenly a lethal trap? No it dosn't. So while not suggesting we ignore any shortcomings let's also not condemn something that actually works perfectly well as long as the operator understands it. Just like every other aspect of aviation you can think of.

I see this inclination to blame the designer as an extension of the Children of the Magenta Mind philosophy or mindset where pilots and commentators seem increasingly unable to deal with matters that the aircraft doesn't resolve or interpret itself, ie a reliance on pavlovian response to specific conditions/warnings without much if any analytical process to back things up where there isn't a Big Red Light that says "XYZ Failure" with the corresponding book-reference requiring "ABC" as response. ECAM is the ultimate example of a system that encourages this mindset, and indeed all but prevents independent action - that damn nearly caused the total loss of a 380 didn't it? Had that crew been merely pavlovian button-pushers they wouldn't have survived.

What we've lost here is something called "Airmanship" (anyone still remember Airmanship?). Knowing and understanding the how and why of your systems, not just which button to press. Knowing your drills and procedures. Thinking through all actions.
Noy relying in the damned autopilot to get you out of every pickle and blaming the manufacturer when it doesn't. We are clearly increasingly becoming reactive operators as oposed to proactive ones.

It isnt, imho, a healthy trend.

Oh brilliant, another “things were better in my day when we could fly manually fly a back beam NDB whilst inverted, without spilling the coffee”, neglecting to mention that there used to be a crash a month. Commercial aviation has moved on, the aircraft have got more advanced and the training and focus is different. It doesn’t mean the pilots are any worse, they’re just using a different skillset and like it or not, aviation is safer as a result. Yes of course you should fully understand the aircraft you’re flying and it’s very much possible to fly a safe G/A in a 737, in the same way that many people very successfully landed a DC-10… there are just gotchas to be aware of. A more modern aircraft such as the 787 would have made that whole scenario considerably easier.

As for the 18 seconds, as others have mentioned that’s not a sign of an incompetent crew… but more likely the conduct of a mini-brief/refresher of what was going to happen, which many training departments encourage if time allows and the ATC instruction isn’t for an immediate G/A.

[meleagertoo]

Zero, you nicely demonstrate my point about failing to assimilate a situation (ie reading what I wrote) by reacting in a pavlovian manner based on an unsupported assumption - viz your first sentence -
As for your last sentence, words almost fail me. 18 seconds to "brief" such a simple manoeuvre? That's one heck of a brief for a procedure that was then completely botched! Look at the timelines in the AAIB report. They don't indicate the actioning of any Boeing procedure I recognise, regardless of how ancient and senile you fondly imagine me to be. But then, I'm not jumping to wild conclusions or blindly following magenta lines, but trying to analyse the facts we all have in front of us.
I reccommend the process.

[zero/zero]

Zero, you nicely demonstrate my point about failing to assimilate a situation (ie reading what I wrote) by reacting in a pavlovian manner based on an unsupported assumption - viz your first sentence -
As for your last sentence, words almost fail me. 18 seconds to "brief" such a simple manoeuvre? That's one heck of a brief for a procedure that was then completely botched! Look at the timelines in the AAIB report. They don't indicate the actioning of any Boeing procedure I recognise, regardless of how ancient and senile you fondly imagine me to be. But then, I'm not jumping to wild conclusions or blindly following magenta lines, but trying to analyse the facts we all have in front of us.
I reccommend the process.

For someone who's all about process, analysis and not jumping to wild conclusions you've made an awful lot of assumptions about the performance of the crew based on what is an interim report and therefore by definition, inconclusive

[safetypee]

Nowadays it is rare for accidents or incidents to have a simple or dominant contribution (a very ‘safe’ industry), thus it necessary to consider a wide range of contributing factors, however unlikely.
More often, particularly with web based forums there is a tendency to reduce situations to black or white positions, and with hindsight conclude that ‘it’ was obvious. These reflect inherent human bias which may not help with the learning required for future operations.

An alternative approach is to consider what the crew did to recover from an unwanted condition, what were the factors which ‘switched on’ a change of plan, what differed from the situation commencing with a discontinued approach to that in recovering from an unintended flight path. Awareness, thoughts, actions, and the many influences on those, from wherever or whenever they originated.

Start by considering the crew as an asset, how they avoided an accident in that type of aircraft, at that airport - context; an everyday activity, operating in varied and changing circumstances. In such situations there may be no simple answer; many opinions, viewpoints, experiences (as in this thread), but not a ‘solution’. At best, aspects which might improve operations, but without assurance of success.

The need is for us to manage the uncertainties in operation, and the uncertainty of our thoughts, before, during, and after an event.

[hec7or]

Erm, having studied the FDM data in Fig 2 of the AAIB report, I would suggest that this is a clear case of failing to realise that the AP didn't engage when the button was pressed.

[zero/zero]

Erm, having studied the FDM data in Fig 2 of the AAIB report, I would suggest that this is a clear case of failing to realise that the AP didn't engage when the button was pressed.

How is that? Which part of Fig 2 shows the AP button being pressed?

Looks to me like AP engaged until the moment the N1 starts to increase, consistent with the pressing of TOGA and the associated disengagement on the AP for a single AP approach. Then it's re-engaged a bit later after the 'recovery' phase is complete

[hec7or]

0/0 I admit it doesn't so on second thoughts and according to the graph, the AP was disengaged at the point of GA and not re-engaged until after the altitude was recovered as indicated in Fig 2, but the A/T remained engaged throughout. Looking at the trim inputs, it can be seen that the aircraft descended at the same time the main trim was commanded APND. This coincides with the increase in airspeed and looks like they may have experienced a somatogravic illusion akin to a pitch up.

Clearly if the crew were using main trim, then they knew the AP was not engaged.

[zero/zero]

0/0 I admit it doesn't so on second thoughts and according to the graph, the AP was disengaged at the point of GA and not re-engaged until after the altitude was recovered as indicated in Fig 2, but the A/T remained engaged throughout. Looking at the trim inputs, it can be seen that the aircraft descended at the same time the main trim was commanded APND. This coincides with the increase in airspeed and looks like they may have experienced a somatogravic illusion akin to a pitch up.

Clearly if the crew were using main trim, then they knew the AP was not engaged.

I’m not an AAIB investigator and don’t pretend to be, but the nose down trim commands are entirely consistent within the airspeed increasing which is exactly where you’d expect them to be. The graphs don’t tell you how big the trims changes were or whether they were appropriate with the speeds, therefore literally impossible to draw any conclusions.

[Chris2303]

Boeing Bashing took on fever pitch over the MAX accidents which once again were caused in the immediate sense by just this, pliots who appeared lacking in their systems knowlege.

It's a bit hard to have system knowledge of an addon that Boeing deliberately kept secret

[Uplinker]

meleagertoo is essentially right that a 737 crew should be able to fly a 737, no matter what the mode and what the situation. My guess - purely a guess - is that probably the F/O was flying and was given an extensive (18 second) briefing about what to do, then unfortunately cocked it up. We've all been there.

As you say, a proper crew should have been able to fly a go-around either manually or with help from the automatics, such as they are. Why this crew - or one of the crew - couldn't do so is the real question, and the answer will be to do with their training and what competency standards were required to pass the SIM in that airline.

'Children of the magenta' often gets trotted out, but unlike that crusty old guy, pilots of today must be able to programme and operate the automatics in today's aircraft and today's skies. There is no shame or weakness associated with using the automatics, and there are plenty of valid reasons for using them. It is not very satisfactory in my opinion to have to keep taking manual control when trying to fly on automatics which are design limited.
There is or used to be a driving licence only for cars with an automatic gearbox, and a person could obtain such a licence and drive only autos. I am not saying that we should ever allow pilots who can only fly with the automatics, but pilots today do need to be able to fly with the automatics.

In a situation like this, you are right, the basic-ness of the aircraft is not directly relevant, but we are naturally going to ponder if its design was a factor in any incident. The 737 is a 1950's electro-mechanical and hydro-mechanical aircraft with some electronics bolted on here and there. Literally - I was amazed to see a box labelled 'fuel quantity computer' bolted to the ceiling above the F/Os head in a 737-300/400 (tel:737-300/400) - why not put it in the avionics bay?

Although not directly relevant, we are bound to ask ourselves why this manufacturer has not updated their aircraft. Why does it still have auto-pilots and auto-thrust that can only do part of the job? Having to have one pilot looking inside to adjust the take-off thrust as you roar down the runway, is not very satisfactory - both pairs of eyes should be looking out at what is going on. I understand it would cost a lot of money to update and also about the commonality required by SouthWest, but the so-called MAX should have been a proper modern update of the 737, with much more capable and properly integrated electronics.

The pitch-power couple is not a particular issue by itself, but having to push and trim forwards to go up during a go-around is quite a design flaw. If it can be designed out with modern electronics, then why not do so? If by some quirk of physics our cars all veered to the right with increasing speed and to the left with decreasing speed, we would all be used to compensating and would no doubt have a manual trim system to compensate. But it is so much nicer that we don't have to do this, and like-wise it is so much nicer not to have to constantly pitch trim in a FBW aircraft. We could all drive around in cars with non-synchromesh gearboxes, having to double de-clutch for every gear change, but it is so much nicer that we don't.

Drivers are able to concentrate more on the driving, and the driving task is much less tiring with modern devices. Ditto aircraft.

[FlightDetent]

personally,
meleagertoo: 2
zero/zero: 0

Is a level-off from an announced, briefed MisAp too much to handle for pilots in 2021? Not the G/A itself, the level-off at the end of it. Boeing shut Classic's production around 1999, IIRC. If the NG is a beast, it should be tamed by now.

This from someone who
- only learned here, 15 years after having changed the type, that his THS trimming techinque was completely wrong resulting in severe overcompensation
- only learned here what speed-tape confusion means and is still scared of his second (yet to come) real encounter with it
- took whole 4 SIM training sessions to get his G/A's polished after the pandemic break - on the mainstay, 'properly designed and pilot-friendly' type.

​​​​I think an honest investigation needs to evaluate the size of the holes in cheese slices too.

[meleagertoo]

It's a bit hard to have system knowledge of an addon that Boeing deliberately kept secret
You illustrate my point perfectly.
There was, of course, no need to, and it is only the pavlovian reaction type pilots that would think it necessary.
Any properly trained pilot would have recognised that the auto-trim was running incorrectly - in other words a trim runaway and remembered (!!!) there is a memory drill to cope with that...

Back to the pitch/thrust couple; a ridiculous suggestion that this is a "design flaw". It is what it is. Is the asymmetric thrust/yaw couple a "design flaw"? No more than oversteer/understeer is a "design flaw" in front and rear wheel drive cars! Is the ideal-handling aircraft a Cessna 337 then? What bothers me is why anyone should aspire to this? Can't pilots cope with differences? Where would Al Haynes be today if that idea had been implemented earler on?

Autopilots and autothrust that "can only do part of the job". They can't taxi or track the runway on take-off! Because they don't need to. Why shouldn't one person have eyes inside on take/off? I thought it was a requirement! There's nothing the least bit odd about that. What is this suddenly urgent need for two of you to be looking outside ignoring the instruments? You can't see anything much in a low viz take-off - so it clearly isn't essential. Any more than it is in IMC.

What is this craving for total automation and totally benign handling? One isn't desireable and the other pointless. Unless you advocate pilotless aircraft of course.

Until this thread I never knew there was a feeling that the 737 had handling flaws - I always thought it was a magnificent aircraft with its own particular characteristics (as does any other aircraft) that were not a problem once they were pointed out. Damn sight nicer to fly manually than an Airbus though - but then maybe we now have a generation of pilots who don't actuallyexpect to be asked to fly any more and expect the autopilot to do everything for them, even when it isn't engaged, a la Airbus...and consider actually flying an aeroplane as a black art. Let's hope they never encounter an aircraft that really does have unpleasant handling design flaws. Let alone a helicopter.

The above merely reinforces my fear that there are too many out there who regard any individualness or character in an aircraft a design flaw and any system that isn't totally automated or that even permits them to mishandle it is a design fault. No pilot error any more, it's all the manufacturer's fault.

[Uplinker]

Well, I agreed with many of your earlier points, but you are wilfully misunderstanding and misrepresenting what I said, so I will leave you to it :D

[safetypee]

Assuming that all crews (any crew) will be able to fly an aircraft, in any situation, etc, does not match the reality of human behaviour; each of us understand situations differently, according to our view at the time, influenced by experiences, knowledge, training.

Concluding a reason for an event without considering the situation as the crew might have done, not appreciating the context without adequate thought, then our views are no better than a guess. Safety is not based on ‘a guess’.
Although safety is not fact, the requirements are based on judgements, including assumptions (although rarely published). A regulator might assume that the tasks (physical, mental) for flying a developed aircraft can be accommodated with a ‘same type rating’.

The 737 has changed; from a conventional swept-wing low-thrust design where controls (and trim) were harmonised with the thrust pitching moment, and a simple (by modern terms) dual channel AP and AT.

The need for Cat 3 auto-land, automatic trim up, an advanced dual FGS requires alternative procedures for GA, for AP/FD, single or dual modes; added complexity. Increased demand on awareness, knowledge and recall, different skills.
More recently increased engine thrust; the manual control system has to cope with a larger pitching moment, but unable to match previous flying qualities exactly - but ‘the crew will manage’ (differences training). And a final mitigation, an afterthought at the end of the checklist ‘If TOGA thrust is not required, then …’; more demand on awareness, decision, action; more complexity in identifying a safety alleviation.

Within this, a critical assumption that the aircraft can be flown with the same basic techniques taught ab-initio; control for pitch, trim to reduce force, a sequential action.
Uplinker identifies the 737 weakness - pitch and trim simultaneously, where any delay adds workload, getting being behind the aircraft.

In this incident, given a need to discontinue an approach: the crew assess the situation, altitude, system status, auto or FD, is the procedure ‘GS’ or ‘discontinued’, and then decide.
A button press on previous variants; now preceded by a multiple choice actions dependent on understanding the situation.
We assume that new technology aids operation (747, 76, 77, 78), but retrofit in an old aircraft changes human activity, with higher mental workload.
Without changing the aircraft or operational situations (money, money, money), all we might change is our thinking - what we assume and why; and with this understanding adjust how we learn from incidents.

[fdr]

1. Assuming that all crews (any crew) will be able to fly an aircraft, in any situation, etc, does not match the reality of human behaviour; each of us understand situations differently, according to our view at the time, influenced by experiences, knowledge, training.


2. The 737 has changed; from a conventional swept-wing low-thrust design where controls (and trim) were harmonised with the thrust pitching moment, and a simple (by modern terms) dual channel AP and AT.

3. The need for Cat 3 auto-land, automatic trim up, an advanced dual FGS requires alternative procedures for GA, for AP/FD, single or dual modes; added complexity. Increased demand on awareness, knowledge and recall, different skills.

4. More recently increased engine thrust; the manual control system has to cope with a larger pitching moment, but unable to match previous flying qualities exactly - but ‘the crew will manage’ (differences training). And a final mitigation, an afterthought at the end of the checklist ‘If TOGA thrust is not required, then …’; more demand on awareness, decision, action; more complexity in identifying a safety alleviation.

5. Within this, a critical assumption that the aircraft can be flown with the same basic techniques taught ab-initio; control for pitch, trim to reduce force, a sequential action.
Uplinker identifies the 737 weakness - pitch and trim simultaneously, where any delay adds workload, getting being behind the aircraft.

1. That would seem to be a reasonable expectation from the self-loading freight, those under the flight path, the company and the regulator...

2. the -100 and -200 series were quite nice performers, In particular ROC. the thrust line was slightly less offset than the CFMI blenders, but doing the maths on the moment arm gives little difference in the resultant, except the thrust level is increased... The -300, -400 and -500 are vanilla designs. The -600, -700, -800 are same again, and the -900 adds some geometry issues that should be a point of concern, but, they all fly the same. Then we got the Max, and essentially the Max is the same except for the high alpha lifting body effect that arises from the cowling. Boeing went for reasons of their own logic that are incomprehensible to me with a repurpose of a system to resolve that particular test point. Aerodynamic modification was an alternative and would have been a benefit in weight and drag as well, but, the repurpose was the solution that was incorporated, and the rest is history.

Neglecting the high alpha Max issue, every B737 behaves the same in principle, in fact, exactly the same as a B747, B744, B757, 767... [The 777 and 787 in normal modes have the PFCs giving a trim reference speed which the plane will settle to, however, they also are both FBW systems that have a phugoid, and that means the pilot may well get to see more sky than expected in a lightweight, double-tap of the TO/GA levers]. The A300-600, and the A310 are spectacular in their ability to get out of sorts in trimming. Perpignan's A320 and many others have shown that the mindset change from normal to direct law in Airbus aircraft can end up being challenging for the driver, while the remainder to use manual THS comes up on the ECAM, history suggests that this little gem in a situation of cognitive overload kind of gets missed, the old "I couldn't hear your go-around call over the sound of the gear up warning horn". Lears, Hawkers, Falcons, Gee whizzes etc (other than the FBW variants (falcon goes C* ways, like the Bus, Boeing likes C*U... and gets a phugoid for its pains) all of these need trimming in Go Arounds, in fact, they usually all need trimming all the time. About the only plane that doesn't need trimming is a Soko Galeb, from stall to 430KIAS you can leave the trim alone and fly by fingertip. At the other end of the scale, a T-28 with the -86R at 1425hp doesn't do anything without trimming in all 3 axis, at the same time, so having 5 hands makes for smooth formation aerobatics... The bottom line, changing the noise level in almost all aircraft needs a trim input to manage the outcome, and aircraft with large speed differences between low and high-speed envelopes, and with large CG ranges will always have a stabilizer that has a high authority compared to the elevator. [neglecting F4's... F8's, A7's, etc... ] The ability to get out of trim is not a unique factor of the B737.

3. Cat 3. itself didn't require the nose up trim input, which is a dual autopilot issue, where it has to achieve a fail passive outcome, If a 3rd AP had been added, 3rd power source, independent instrument transfers etc... then fail-operational would have been possible which does not involve a nose-up trim bias input. B747s, triple or dual autoland could be conducted, Triple avoided the out of trim if a manual landing was made with a late AP disconnect. The B737 dual AP has to be fail passive, hence the trim bias.

4. for the variants, the moment arm change for the tail offsets the thrust change on most heavier models. the shorter body big blender types will usually want some additional thumb flickering for a full thrust GA, but all need trim input.

5. Attempting a GA in an A300 or 310 by a sequential application of elevator then trim is going to be.a great airshow. bring popcorn. Same with a light B763 with big GEs or PWs. Doing the same with a T28B or D will give a neat tight barrel roll, [ well, things get fun]. Hit the TOGA 2 times on a light B744 and try to fly with elevators only and you will get sore arms, and still see lots of sky. MD11's were fun too. A large thrust change in most jets requires prompt elevator AND stabilizer trim inputs, commensurate with the rate of change of the moment and the magnitude of the moment change. nothing magical in that.

Swept wing high-speed transports have a high ratio of Vmo/Vs, fairly large Cp changes from transonic Mach effects, and wide CG ranges, and almost invariably have a limited elevator authority relative to stabilizer authority. (Boeing always used to desire a plane that could be flown by the driver using just the control column in an engine failure, which while a wonderful sentiment is itself a lousy real-world proposition.... just as if the thrust couple changes it necessitates a trim change (trim=stab...) a thrust asymmetry requires an appropriate trim change for the yaw, that is a rudder input. [countering roll only with aileron in almost all modern and not so modern jet transports results in the potential for an increase in stall speed where yaw has not been controlled].). going right off track, Cp rearwards shift with increasing mach No is a relatively gentle affair, however, the term tuck certainly comes to have a meaning when a shock develops on the lower surface of a wing, which results in a rapid change in Cl, and a reduction in the flight path angle from the loss of lift.


The B737 is just a plane, it is not my favorite Boeing, it has sloppy ailerons, and the dual-acting single servo rudder control valve was about as bad as the. concept of MDD's fail-safe stab screw jack design, or airbus's A300 vertical stab secondary structure failsafe design... . but push and pull is conventional, even with the MAX where the crew is let into the secret of the MCAS system being fitted.

Between C* and C*U, the C* flies nicer when it works. The C*U as Boeing implemented it has the need for the driver to use the trim switches to re-index "U", and so the reversion from normal to degraded laws is not a difficulty. The Airbus obviates the trim wheel use by the driver in normal flight, but then when stuff gets rowdy the crew has to have the presence of mind to use the stab trim (THS) lest they conduct an impromptu Aresti series.

[Denti]

3. Cat 3. itself didn't require the nose up trim input, which is a dual autopilot issue, where it has to achieve a fail passive outcome, If a 3rd AP had been added, 3rd power source, independent instrument transfers etc... then fail-operational would have been possible which does not involve a nose-up trim bias input. B747s, triple or dual autoland could be conducted, Triple avoided the out of trim if a manual landing was made with a late AP disconnect. The B737 dual AP has to be fail passive, hence the trim bias.
The 737 is perfectly capable of fail operational CAT IIIb operation. It is just that most carriers do not want to spend the extra money on the maintenance for the rudder servo that is fitted additionally. The autoflight system is inherently capable of CAT IIIb out of the box though, it is just not activated for most carriers. And the CAT IIIb fail operational airplanes still have the same trim up thing going on, both in dual engine and one engine out autolands (yes, those are approved as well if one has the fail operational version, flaps 30 for landing as well).

[fdr]

The 737 is perfectly capable of fail operational CAT IIIb operation. It is just that most carriers do not want to spend the extra money on the maintenance for the rudder servo that is fitted additionally. The autoflight system is inherently capable of CAT IIIb out of the box though, it is just not activated for most carriers. And the CAT IIIb fail-operational airplanes still have the same trim up thing going on, both in dual engine and one engine out autolands (yes, those are approved as well if one has the fail-operational version, flaps 30 for landing as well).

That is quite correct. It has a curious 2 APLT 3 IRS reference architecture that provides for the fail-operational capability. With 2 APLT and without the option, it is fail passive with 2 APLTs. There is a note in the FCTM for the plane that says, for single-channel ops, refer to limitations section of the AFM, and in most cases that requires a single channel to be disconnected no lower than 50ft AGL, for FAA and 158ft AGL for EASA....

[safetypee]

fdr, thank you for the additional information and safety views.

Reflecting on the 737; yes another aircraft. However, not ‘just another’ in comparison with others.
What is the level of change in the 737 in comparison with other types, i.e. % increase in thrust, aircraft mass, … Also, noting the increased tail-plane area, adapted wing shape and profile.
These changes are not seen as cause, but contributing factors influencing normal operations, which over time have increased the complexity of operations.

Re passengers expectation - the same for the crew, but safety is judged on outcome, where the gap between the extremes - expectation and reality, has to be managed by the crew.
Crews create safety; the crew in this incident recovered the situation, a safe outcome.
There have been many opinions as to how the event was initiated, but few (none) about the recovery - how was this achieved.

P.S. only flew the T28 at half boost (3 flts). A7 LoC, hands off everything (3 flts)

[Denti]

That is quite correct. It has a curious 2 APLT 3 IRS reference architecture that provides for the fail-operational capability. With 2 APLT and without the option, it is fail passive with 2 APLTs. There is a note in the FCTM for the plane that says, for single-channel ops, refer to limitations section of the AFM, and in most cases that requires a single channel to be disconnected no lower than 50ft AGL, for FAA and 158ft AGL for EASA....

Yup, pretty much the same as the A320, which uses a two autopilot and three full IRS (737 uses two full IRS and one ARS) system. The three autopilot thing had to be used in the past, technology has evolved and two are enough now. The 737 fail operational system is actually more modern than the A320 one, by around 15 years. Still, the operation is in many ways hampered by the insistence on single channel approaches which is neither necessary nor clever. Although, to be honest it wouldn't have made a difference in the original case of this thread.

[lederhosen]

To answer safetypee, I suspect the recovery was initiated when the crew belatedly realised that the autopilot was not engaged. The report will confirm who was flying. However I am familiar with Aberdeen and whilst not difficult it is relatively short and as already pointed out busy mainly with North Sea helicopter traffic. My money would be on the captain giving the easy leg to Palma to the FO who had not done a lot of flying recently with the captain flying to Aberdeen. The report says he had a lot of hours but less than 10% of it was on the 737. He had also been off flying for an extended covid related period, only returning the previous month. This is pure speculation. But I could easily imagine a scenario where the captain pressed TOGA (reverting perhaps to old habits) and thinking the autopilot would fly the go-around and the two of them became task saturated cleaning up the airplane and not overspeeding the flaps with nobody flying. As the sink rate and speed increased someone (quite probably the captain) took control of the aircraft. It is salutary how quickly a high rate of descent built up and the time it took them to react. It seems the authorities were also alarmed that this could be indicative of wider problems with out of practice crews.

[FlightDetent]

On the NG when the yoke elec. rocker sw is activated, does the trim wheel spin always at the same speed? Asking here to learn whether or not there are different 'sensitivity' settings among the various flap positions - seem to recall from somewhere it actaully was configuration dependent.

Two more, if I may: What is the standard retraction sequence for flaps/slats after G/A, i.e. which of the intermediate postions are used: 30-15-5-0 ? When cleaning up, does any of the configuration changes introduce a N.D. moment?

[switch_on_lofty]

Manual electric trimming rate is fast with any flap extended and slow with flaps up.
Go around is initiated by calling "go around, flap 15" whether at flaps 30 or 40. You're correct that the normal cleaning up sequence is 30/40 - 15 - 5 - Up.
In terms of trim change it's hard to say because a lot of other things are going on at the same time; probably thrust change, increasing speed, speed trim system and levelling off. I'd summarise by saying that no flap changes have an immediate and large pitch effect but the effects are masked by drag/profile changes anyway.

[172_driver]

The 737 fail operational system is actually more modern than the A320 one, by around 15 years. Still, the operation is in many ways hampered by the insistence on single channel approaches which is neither necessary nor clever.

Ever had the AP disconnect close to DH on an actual CATIII appr? I have. There is a certain startle effect to it and still in the milk, there isn't much visual to go by. It's essential that the crew has the motor skills quick at hand to perform a manual go-around. I again boast that a single channel approach should be standard if the weather is reasonable, as it might be the only real (and I mean real, not simulated) practice you may get at one. One day you might need it.

[Denti]

Ever had the AP disconnect close to DH on an actual CATIII appr? I have. There is a certain startle effect to it and still in the milk, there isn't much visual to go by. It's essential that the crew has the motor skills quick at hand to perform a manual go-around. I again boast that a single channel approach should be standard if the weather is reasonable, as it might be the only real (and I mean real, not simulated) practice you may get at one. One day you might need it.

Of course there is a startle effect, which is why good training tries to focus on startle effects and how to get out of that startle, by the way, the startle effect hits you the same on more stable aircraft like the A320, probably even more so as the alert height is only half as high. That said, apparently there is quite a big risk for pilots to really mishandle manual go arounds, many people have paid the ultimate price for that. I would argue that a dual channel approach would have been better in that case: click the TOGA button, let the automatics do the rest, wrap your head around the somatogravic illusion, and once settled one can use whatever he wants including manual flight. Granted, it wouldn't have helped the TUI guys, but it certainly would have helped the FlyDubai crew.

[hans brinker]

Ever had the AP disconnect close to DH on an actual CATIII appr? I have. There is a certain startle effect to it and still in the milk, there isn't much visual to go by. It's essential that the crew has the motor skills quick at hand to perform a manual go-around. I again boast that a single channel approach should be standard if the weather is reasonable, as it might be the only real (and I mean real, not simulated) practice you may get at one. One day you might need it.

This discussion will always be there. Automation makes aviation safer, automation dependency makes it less safe. In order to avoid automation dependency, you have to avoid automation. If you only avoid automation when there is no other stressors, it will be less safe when the automation fails when there is..... As a pilot who avoids automation more than most on my fleet, I think if we tried avoiding the kind of accidents like what happened on the MAX by having everyone hand fly to and from TOC/TOD, we would see much better flying skills on average, but more accidents as well.

[Busbuoy]

And if employers had to hire from the pool of people capable of hand flying to and from the cruise they would have to pay a lot more, and they will never let that happen again.

[172_driver]

This discussion will always be there. Automation makes aviation safer, automation dependency makes less safe.

Well put, and I understand the catch. What would be an ideal balance is that aircrew at least aspire to better themselves, try the water, on every flight. Start with visual conditions at a calm airport. Next time try it when the ceiling is lower. Or at a busier airport with headings, altitude, speed instructions. If you're approaching overload, engage! Shying away from it will make you less ready the day you need it. And prone to suffer from illusions more easily. It's been done to death before, I know.

[Alrosa]

Well put, and I understand the catch. What would be an ideal balance is that aircrew at least aspire to better themselves, try the water, on every flight. Start with visual conditions at a calm airport. Next time try it when the ceiling is lower. Or at a busier airport with headings, altitude, speed instructions. If you're approaching overload, engage! Shying away from it will make you less ready the day you need it. And prone to suffer from illusions more easily. It's been done to death before, I know.

Couldn’t agree more. Hopefully the person sitting opposite is willing to do their bit.