B737 Classic- A/T de-select speed

[Wizofoz]

After a long break, I'm currently re-qualifying on the B737-3/400.

Back-in-the-day with another operator, we had an often used procedure of A/P disconnect, but rather than fully disconnecting the A/T, we would de-select speed, leaving the A/T in Armed mode, the theory being it gave low-speed protection.

Later I became aware that Boeing recommended against this procedure, as they thought it risked un-commanded A/T movement at critical times(such as flare/landing).

Lo and behold, my new company also uses this procedure.

Does anyone have a link to an actual Boeing guidance source recommending against this? I'd just like to diplomatically show my management that the manufacturer doesn't like this procedure.

[vapilot2004]

I would check your training manuals, Wizofoz for the reference which states that Boeing recommends manual flight with A/T during takeoff and climb out only. There have been more than a few SBs on the A/T system for the Classic as well.

I have been told that aircraft energy management is the reason behind the recommendation. The A/T system is designed to work in concert with the CMD functions of the A/P. Manual pitch changes can lead to an undesirable (high) energy state during landing when A/T is engaged, particularly during gusting conditions.

You can also end up needle chasing (airspeed) when you get out of sync with the A/T control loop and a sort of hysteresis can occur, a condition that is extremely rare with the A/P engaged.

[Wizofoz]

Thanks VA- But not sure we are on the same page. De-selecting the A/T puts the system into ARM mode, meaning it does not control the thrust UNLESS it reverts to SPD due to "ALPHA FLOOR" (Which is really min-speed)

I do have the reference in the FCTM which seems to not recommend using ARM mode, and will raise this with management- just wondered if anyone else had a Boeing reference that discouraged the practice.

[Chesty Morgan]

Not a Boeing reference but our manuals (which I can't link to sadly) corroborate yours.

Still, it's a practice widely used and not discouraged. Go figure.

[vapilot2004]

"I see", said the blind man. Apologies, Wiz. I think my eyes may have glazed over.

The change in A/T ARM mode guidance was fairly recent, like in the past decade or thereabouts. Wish I could be more specific, but that might be a good place to start looking. I am fairly sure avoiding mode confusion was one of the reasons behind the change.

[galdian]

AFAIK Boeing only recommended NOT deselecting as often 737 was first jet type for turboprop operators so KISS...all manual or all auto.

If an operator had proper training on the benefits (alpha floor and G/A thrust with TOGA) and the negatives/times should be disconnected completely (gusty/swirly wind conditions with large airspeed fluctuations) then Boeing lets you fill your boots.

It IS only a recommendation, not a requirement and the FCTM only a reference guide - although a bloody good one and grossly under used IMHO - as opposed to company SOP's, whether stated or implied.

Personally additional protection (safety) for (say) 95% of the time Vs less protection (safety) for 100% of the time....know which I'd choose.

Cheers.

[A37575]

AFAIK Boeing only recommended NOT deselecting as often 737 was first jet type for turboprop operators so KISS...all manual or all auto.


I was unable to locate that additional "advice" you quote (?) from the Boeing 737 FCTM about "as often 737 was first jet type for turboprop operators so KISS..all manual or all auto". Someone has been pulling your leg, methinks:ok:

The FCTM does state: The autothrottle ARM mode is normally not recommended because its function can be confusing. The primary feature the autothrottle ARM mode provides is minimum speed protection in the event the aircraft slows to minimum maneuver speed . Other features normally associated with the autothrottle, such as gust protection, are not provided. The autothrottle ARM mode should not be used with non-Normal Checklists. Some malfunction that affect maneuver speeds cause the autothrottle to maintain a speed above approach speed.

If your company chief pilot wants you to use the ARM mode for approach and landing, then he would be wise to contact the manufacturer and request their advice. In turn the manufacturer will likely say read your FCTM or maybe even "fill your boots" and quietly advise your insurer on the side:E

[galdian]

Well....all those decades ago as a new, humble F/O on the 737 I didn't think it was my position to tell my leaders they were talking :mad:

The explanation made sense - still does - and de-selecting appeared to work well in the airline and has worked fine for me for over 25 years.

Pleased to see the words "normally not" in your FCTM statement, makes one wonder when Boeing WOULD recommend...maybe if proper ground school/SIM/line ops training and demonstration was completed, sadly appears either modern FLT OPS departments are incapable of including such training or maybe it's just too technical a concept for the modern pilot to understand?

Now back in MY day....

Anyhow whatever the wisest of the wise in your FLY DEPT bang on about will have to suffice.
Cheers. :ok:

[Derfred]

Fascinating. I was taught to have an occasional glance at my airspeed during approach. I must be getting old.

Can you put the autopilot into ARM mode so it will flare if you don't get around to it?

In all seriousness, I can think of a couple of recent 777 crashes caused by the pilot expecting the autothrottle to move and it didn't... Maybe mixed automation isn't all it's cracked up to be.

[galdian]

Is it mixed automation?

Maybe it's more back-up/redundancy which improves....safety.

In ARM mode - apart from excessive wind gusts when consider disconnecting as discussed - the A/T will only engage if the pilot :mad: up and allows speed to decrease towards the stall.
And that's only after it flashes a warning for a while to alert the pilot of what the A/T MIGHT have to do if the pilot doesn't correct THEIR error.

A system that tries to prevent the aircraft from stalling if the pilot gets distracted - mighty dangerous piece of equipment, better not use that!

At any stage - regardless of what systems are/aren't being used - the pilot has an obligation to make sure what should be happening IS happening, if not then MAKE it happen.
If pilots decide NOT to make things happen....well that's a whole other thing.

[Centaurus]

and de-selecting appeared to work well in the airline and has worked fine for me for over 25 years.


And some crooks have been breaking into houses for 25 years and not being caught. That doesn't mean to say it is legal and safe:E

Seriously, though..one sequence where selecting ARM on the A/T for manual approach and landing can bite you in the bum, is the All Flaps Up Landing.
This might be one of the sequences cautioned by Boeing when it says:
The autothrottle ARM mode is normally not recommended because its function can be confusing

Thankfully, few pilots have ever been faced with an All Flaps Up landing in the real aircraft in their career. Me for one. But in the simulator I have observed on many occasions where a manually flown all flaps up landing is conducted with ARM mode selected, the throttles started to open un-commanded seconds before the flare as speed is just on VREF. Of course a wise pilot will firmly hold the throttles to prevent this happening - but not all of us are wise and the creeping open of the throttles at the flare might be unexpected.

Which is probably why the FCTM adds: The autothrottle ARM mode should not be used with non-Normal Checklists. Some malfunction that affect maneuver speeds cause the autothrottle to maintain a speed above approach speed

[galdian]

More than fair, Centaurus!

If your flying around constantly referencing the NNC then A/T disconnected may indeed be the way to go.

HOWEVER....if you're flying around constantly referencing the NNC there are probably bigger questions you should be asking about the professional airline operation, the A/T question should pale into insignificance!

All good fun

[Derfred]

Is it mixed automation?

I'm just not so sure that having it in ARM as "protection" is really as safe as you make out.

The pilot "knows" it's in ARM, so may be expecting it's protection. That could be more dangerous than it's worth. I consider it bordering on mixed automation, which has already killed two 777's.

Sorry, I'm a big fan of "all on" or "all off". At least you know where you stand if the situation suddenly goes pear-shaped.

[RAT 5]

Sorry, I'm a big fan of "all on" or "all off". At least you know where you stand if the situation suddenly goes pear-shaped.

That's the Boeing idea, and generally it works fine. There's always a but. I used to fly B733 in 90's and in an RA it was A/P off. No mention of A/T. I moved to B738 and it was all off. I then checked back with the up to date B733 book and it had changed too. When I asked a Boeing pilot, he repeated the Boeing mantra. I've seen too many low speed events with high level RA's as PF forgot to reconnect A/T and PM was still breathing fast. Low interest to those comments.
On finals I'm not a fan of anything that reduces PF's involvement with manually manoeuvring the a/c and dulls the loop of PF-aircraft hand- eye coordination. Derfred has a good point. Don't encourage PF to reduce their scan of what's going on. It is not a playstation and needs positive control of all parameters, and speed is one the more critical. The pilot's brain must be well ahead of the a/c and therefore sharp & alert.
How many operators make A/T mandatory for LVO approaches? Some do. It's not a Boeing thing, but if airlines begin not to trust PF's skill in maintaining speed in manual flight, they might carry that across to auto flight, unnecessarily technically.
There is also the case to consider of not allowing pilots to operate under conditions that may be allowed and expected under MEL guidance. A u/s A/T is not a NoGo, and it is a single system i.e. no back up. If crews are not used to it being off, but are then expected to operate a whole flight in manual thrust, it might prove a little too much for some. We've seen the consequences of mollycoddling pilots in too much automation. Considering my early days in B732 and last days in B757/767/738 I've seen a wonderful evolution of automatics and enjoyed many of the improvements on offer; and never felt I'd lost any control. What did make me feel emaciated was mollycoddling SOP's.

[galdian]

We all have stories where "things happen" with various systems engaged.

A go around in min wx by a "new" old esteemed Captain where he'd disconnected A/T, applied way too little power and ended up with speed decreasing with a SINK rate of 800fpm below 1000ft AGL before the (fortunately) experienced F/O was able to "assist" the Captain in recovering without having to take over and all the "face" issues that would have caused.

One wonders what would have happened with a newer, more socially correct F/O.

He pressed TOGA correctly, A/T in ARM would have been a non event instead of far too close to a smoking hole.

RAT 5 you have me confused - I think you're saying A/P off and A/T on which I totally agree is no go - but A/T in ARM means manual thrust/speed control by the PF up to the point where the PF has stuffed it up, lost attention/scan, speed decays....and the ARM mode kicks in to 1. advise the PF he's stuffed up then 2. prevent the aircraft from stalling.

ARM is not an active mode - unless the PF stuffs up and tries to stall the aircraft.
PF doesn't stuff up - the ARM mode has no need to do anything, will not interrupt manual thrust operation and will click itself off 2 seconds after touchdown.

Pure Boeing magic! :ok:

[RAT 5]

Galdian: apologies. I confused myself. Indeed, as you say. Deep in my memory, and playing hide & seek, is an item from B757/767 days where there was the ARM function in daily use. (I stand to be corrected). The reason I think I remember this is when I transferred companies onto B733 there were other B757 pilots who experimented with the 'hidden' ARM function and found it seemed to work. In that embryonic airline they discussed its use within the training dept only to be rebuffed as a no no. It was taken as a case of the people being spoken to were less experienced that those speaking and didn't understand the question. So the Ops guys went to Boeing and received the standard Boeing 'not recommended' and thus it became a DO NOT. On the outside it seemed the B757 AFDS was the same as B733, but we all know there are gremlins lurking behind the facia deep in the electronic bowls of the E&E bay. What seemed to be happening in Seattle was that the B737 fleet development dept operated separately to the very similar looking B757 dept. They introduced some similarities, perhaps in the hope of common type ratings, but not everything. (who ever thought that using the flap lever to drive the speed bug on GA was good idea. Duh!) I still think B737 ARM is a grey area. It might well work and work well, but I wonder if there are certification problems that are better left unopened? As a result, 'not recommended' is interpreted as 'not allowed' by the many, but not all. If some XAA's approve it what argument do those who deny it use?

[Centaurus]

As a result, 'not recommended' is interpreted as 'not allowed' by the many, but not all. If some XAA's approve it what argument do those who deny it use?

Interesting points made Rat 5. It becomes a problem when type ratings are undertaken in the simulator and students are actively encouraged to thoroughly study the various FCOM and FCTM.
Depends too on the simulator instructor's previous background. So one instructor says its OK to de-select speed and leave the AT armed. Another says just do what the Boeing books say. There are no crashes either way so the student decides on the safety of contraception - i.e.de-selects speed just in case so he won't stall on final.

There are pilots who get twitchy without the crutch of some sort of automation. I have seen them flying visual circuits in the simulator with the flight director on, using split cues - that is pitch bar showing altitude hold and heading mode off. The fact that the altimeter is the primary pitch "bar" for level flight escapes them. Thus their addiction to the automatics need another aid - the FD pitch bar. I could be naïve, but IMHO Boeing have carefully considered the pros and cons via extensive research before committing recommended techniques to paper. That presumably includes switching off the autothrottle system when manually flying on final approach in a 737

[Denti]

Depends too on the simulator instructor's previous background. So one instructor says its OK to de-select speed and leave the AT armed. Another says just do what the Boeing books say. There are no crashes either way so the student decides on the safety of contraception - i.e.de-selects speed just in case so he won't stall on final.

Conflicting information during training is of course a big no no. I flew the 737 classic and NG in airlines that used the ARM function, consequently the FCOM was adapted as well as the training to reflect that, the boeing warning was not contained in the FCOM anymore and every trainer advised us to use the ARM mode, and we were trained in its behaviour. Worked like a charm and had some positive and negative sides. In weather with strong gusts we tended to switch the AT off completely as intermittend autothrust activation above bug speed happened.

We didn't see it as a crutch, but rather as a safety device, and funnily enough boeing did agree on its frequent audits.

[stator vane]

Just a personal perspective. I never saw any useful benefit of that use of A/T. I always preferred either an A/T ON or OFF control of the engines, also resting my hands on the thrust levers to tactfully confirm they both are doing what I want them to do. Back on some prehistoric 300s we saw that one thrust lever would come up whilst the other stayed at idle. A real surprize that prompted me to leave my hands on the thrust levers almost all the time I was expecting them to be moving either way. On another but slightly related subject, use of CWS was very useful at times, but Boeing now says NO to that and our 800s will not go into that mode via overpowering the control wheel or column.

[galdian]

stator vane

with A/T in ARM you have as you wish...manual thrust, you can push and pull to your hearts content, the A/T will have absolutely no input until the PF tries to stall the aircraft...like on a min wx circuit at night when it's all happening, get a bit distracted and the airspeed drops off,

And that's after it's done you the courtesy of saying "hey fella you're going to force me to do something unless you fix YOUR error".

ARM is NOT an active mode, it is a back up and a redundancy so, by definition, improves safety.
And as soon as I mention the word "safety" I win, end of story! ;)

Denti - nice explanation, thanks.

I think an important point to emphasise is that you should be trained and the use (and times when NOT used) demonstrated in the SIM and discussed at suitable times during line training, get the impression it's not being trained then the wisest of the wise in Flight OPs say "why are we having problems"??

[Wizofoz]

Always nice to start a thread that get's an interesting discussion going!

Definitely agree the practice has it's advantages IF it is applied with rigorous caveats (Not used with non-normal configs, not used in gusty conditions etc.).

That is where the problem can lie, and it largely comes down to the ethos of the operator- strictly prescriptive or open to different techniques.

Spent the last decade with the former, and now am working for the latter- and probably need to adjust my thinking accordingly!

[ussatlantis]

The airliner I flew for did not have this procedure. I kind of love it though...not for low speed protection but for go around.

I always had my hands on throttle when flying manually and found my self overriding it occasionally when I did not like what it was doing.

[Mansfield]

This is a terrific discussion for people like me, newly typed in the NG at this late stage of life (don’t get me started…). I really like the airplane, brings back enough 727 memories to last me and is so antithetical to the Airbus that I am practically gleeful. However, our training department, as is the case in many settings, lives in a stovepipe, enjoying life without the slightest notion of how any airplanes actually work other than their own. For us more senior, less flexible types, coming off years of 767/757/MD80 operations, the nuance of the 737 has been challenging. Wouldn’t hurt if the manuals were slightly more informative, but maybe that’s just me…

I started out getting chastised for pickling off the A/T at 100 feet…”we don’t do that…” “Do what?” “Disengage the autothrottle…” “What? Why?” “We just don’t…”

We had a bit of discussion about the “technique” of deselecting speed and leaving the autothrottle in ARM. As others have pointed out, this is frowned upon. Interestingly enough, if I understand it correctly, that is exactly what will happen at VNAV disconnect, which will occur at the end of the approach. This could be some distance from the runway. In any event, VNAV disconnect will leave the A/T in the ARM mode exactly as would deselecting the speed mode. I think.

If I’ve got this right, not only does the A/T arm mode give you alpha floor protection, but it also is armed for go-around (below 2000 feet, etc., etc.) Hence the wonderful ability to get autothrust during the FD-only go-around following the VNAV disconnect with no runway in sight.

The only drawback I could see to deselecting speed mode is that one has to reach up and push the correct button without smashing your cheek on the HUD combiner while peering around trying to find said button. On the other hand, as I was told I would, I’m not sure I see any real advantage over just leaving the A/T active and manually overriding it as needed.

RAT 5, "who ever thought that using the flap lever to drive the speed bug on GA was good idea. Duh!" is the best explanation of that gimmick that I've heard. Apparently I was supposed to figure that out by osmosis. :ugh:

[galdian]

Just make sure you have correct terminology and thought process:

A/P engaged and A/T connected...OK
A/P disengaged and A/T disconnected...OK

A/P disengaged and A/T switch selected on...NO
A/P disengaged and A/T DESELECTED (so in ARM)...WHY NOT (reversion mode in case PF stuffs up, for safety)

Yes G/A thrust (as well as FD guidance) provided after TOGA selected. 1 click 1000-2000fpm ROC, 2 clicks full G/A thrust...lots of grunt and lots of fun!

[Centaurus]

2 clicks full G/A thrust...lots of grunt and lots of fun!

"Lots of fun" it may be but there is a serious side to it. To prove that point, one of the first demonstrations to pilots undergoing a 737 type rating is to demonstrate the strong pitch up that occurs when high thrust is used on go-around and the subsequent need to keep that pitch up under control before it bites you. Forget the FD as this demo should be raw data so that the pilot can see the artificial horizon display un-encumbered by wavering FD needles

The demo is simple. Have the student manually approach in the full flap landing configuration. At 100 feet direct him to go around with full power while simultaneously asking him to remove his hands clear of the control wheel. Ask him to observe the rate of pitch up and pitch up angle. Freeze the simulator as the pitch up passes 45 degrees and point out that the aircraft will keeping on pitching up under full power until it stalls.

If ever a demonstration convinces the student the need to actively control the pitch attitude during a low altitude full power go-around in the 737, this demo should do the trick. A picture is worth a thousand words.

[B737C525]

The most powerful flying control in the 737 is the stabilizer. Why that most powerful control is hidden behind two little switches on the yoke is Boeing’s business, but the fact that it is, has claimed lives (TK1951 et al). Conversely, automating that control in the way the French did, and making it transparent to the crew, is also deadly (AF447).

The stab is sized to cope with the full power go-around under the least favourable circumstances, down to approximately Vref-5 KIAS. Below that speed, control is not assured, and a reduction of thrust is necessary.

Each time I look at the tail I compare the size and range of movement of the elevators, and the size and range of movement of the stab, and remind myself of that.

Regarding the OP’s question: The autothrottle was not designed to ‘wake up’ and provide low speed protection during landing. The fact that it does something reminiscent of the envelope protection of other designs doesn’t make it fit for that purpose, as the manufacturer cautions. I’m not a fan of blind acceptance of manufacturer’s guidance, but in this case, I stick with what Boeing say.

[galdian]

...but it will "wake up" on a dark night, scud-running min wx approach at 600 ft if the PF's scan rate fails and he allows the speed to approach the stall

...and it will provide G/A thrust upon TOGA for a go-around (which of course the PF will back up anyway and not forget!)

...and it wont have any effect on an approach unless, once again, the PF's scan rate fails and he allows the speed to approach the stall

Looks like a lot of good stuff for redundancy/safety.

Cheers. :ok:

[Wizofoz]

Wellll......

Gald, it's interesting that you mentioned it not being a good idea to have A/P off, A/T on- I agree entirely, but you might recall that this was ALSO a practice used at our erstwhile employer (in fact, BOTH our previous mutual employers!)

I took that with me to Blighty, and it almost bit me in the arse, demonstrating it to an F/O onto a short runway on a Greek Island, the supposed advantage being you could carry the minimum 5kt additive regardless of wind.

For reasons unknown, the A/T decided not to go into "retard" at the appropriate RAD ALT, instead INCREASING thrust to try and keep VAPP.

One of them "the time I probably should have gone-round but didn't" moments!!!!

My point being, the old automatics in the classic-slug can be a bit unpredictable, so I'm still a little wary of having them on in ANY mode near the ground.

[RAT 5]

There must be someone with a conduit into the Boeing B737 development department who can ask 'the horse's mouth' and relay the words of wisdom to us all: the technical reasons, not just a "we haven't thought about it too much so think it's not for best......"
There are other 'not recommended' techniques mentioned which I've not fully understood the thinking behind. I've tried their use in the sim and they worked fine and made life easier in an engine failure case; and I failed, via the in-house tech dept, to find out the reasoning. The tech dept wasn't interested and so didn't pursue the issue, just forbade it.

[galdian]

Welll...right back at ya! :ok:

Our second mutual had many quirks for many reasons in the overall operation, the first somewhat more SOP orientated....with relatively good to great training IMHO.

I honestly have no recollection of split active auto modes - ie A/P off but A/T on, that's always a no no and only takes seeing someone in the SIM pushing and pulling the yoke (manual flight) and the A/T (auto) advancing and retarding in response to the pitch changes and the whole thing just getting worse...and worse...very untidy.

It was always:
- "deselect speed" (A/T in ARM/redundancy mode so NOT active), then
- "disconnect autopilot".

End up with Manual flight, Manual thrust, Alpha floor protection available and Go Around thrust available via TOGA.

Beautiful....as only Boeing can!

Cheers.

PS: once we've sorted A/T ARM we can talk about the magnificence that is CWS! :D

[sonicbum]

...but it will "wake up" on a dark night, scud-running min wx approach at 600 ft if the PF's scan rate fails and he allows the speed to approach the stall

...and it will provide G/A thrust upon TOGA for a go-around (which of course the PF will back up anyway and not forget!)

...and it wont have any effect on an approach unless, once again, the PF's scan rate fails and he allows the speed to approach the stall

Looks like a lot of good stuff for redundancy/safety.

Cheers. :ok:

You’ve mentioned twice what if the PF scan rate fails, so You assume that on a dark night the PM is already sleeping or on WhatsApp ?
4 eyes should be sufficient to have all these wonderful protections for about 40 seconds (let’s assume You disconnect the automatics at 500ft or so).

[B737C525]

Alpha floor protection available

No. There is no alpha floor on the 737. If you think there is, and you think that leaving the autothrottle armed provides it, you are sorely mistaken.

[Chesty Morgan]

No, it's low speed protection. To all intents and purposes it's the same thing.

[galdian]

B737C525

Quite correct, my apologies for incorrect terminology: Minimum Speed Reversion is what I should have written.

sonicbum

Well in some operations/cultures in those 40 seconds the PM will probably have to set a missed approach heading, set a missed approach altitude, de-select and re-select Flight Directors whilst (maybe) responding to various other PF directives and random ATC comms.
His actions may not be limited to solely monitoring instruments and calling deviations.
Easy enough for both PF and PM to get distracted at inconvenient times, another case of "beautiful SIM script" Vs "all hell breaking loose" at times in the real world.

Surely a bit of redundancy isn't going to DECREASE safety?

[sonicbum]

B737C525

Quite correct, my apologies for incorrect terminology: Minimum Speed Reversion is what I should have written.

sonicbum

Well in some operations/cultures in those 40 seconds the PM will probably have to set a missed approach heading, set a missed approach altitude, de-select and re-select Flight Directors whilst (maybe) responding to various other PF directives and random ATC comms.
His actions may not be limited to solely monitoring instruments and calling deviations.
Easy enough for both PF and PM to get distracted at inconvenient times, another case of "beautiful SIM script" Vs "all hell breaking loose" at times in the real world.

Surely a bit of redundancy isn't going to DECREASE safety?

Galdian, if the PM ends up being so busy and the Work Load so high at the latest stages of the approach I believe there are other issues that need to be addressed rather than debating the use of A/T ARM or Off. Redundancy is going to increase safety IF it does comply with the manufacturer’s guidelines.

[Mansfield]

This discussion began with a question about the 737-3/400. I can only speak to the 737-800, although I believe there is great similarity in the autoflight systems.

Regarding alpha floor protection, the Boeing FCOM states: “alpha floor protection automatically engages the A/T when armed.” Can’t be much plainer than that. That said, this reminds me a bit of my insurance company claiming that they will cover the damage to a rental car. Haven’t tested it yet to see what actually happens when I crash the rental car…and I have no intention of testing the alpha floor.

Minimum speed reversion appears to be a mix of pitch and power controls to return the airplane to a speed above whatever the minimum speed is. Alpha floor is not defined anywhere except in the above statement, but appears to only an A/T mode.

Regarding the split in automation, my strong opinion coming off the 767/757 and MD80 before that was autopilot off, autothrottles off. I still prefer to fly that way for all of the reasons cited above, and have often invoked Boeing’s admonition on this topic. However, in 2006 Boeing’s Captain Mike Bryan produced a presentation at their annual operators’ forum which questioned that premise. He showed quite a bit of flight test data indicating a much better and more precise speed control in gusty crosswinds, etc. with autothrottles on during manual flight.

Unfortunately, in 2008, my access to Boeing’s operator forums and associated presentations was summarily cut off (airline folded). So I don’t know what else they may have done with this issue…perhaps some of the other graybeards on this forum might. BUT…regardless of personal tastes and historical recommendations, I am now required to manually fly a CAT III approach using the HUD…with the autothrottles on through the touchdown. So it’s a little hard to support the notion that there is something inherently unsafe about splitting the automation.

Regarding the retard action in the flare, I certainly can see Wizofoz’s experience on the Greek island happening. We were drilled on the importance of manually closing the throttles during the flare…except during the aforementioned CAT III…even if they are active. The primary issue here seems to be the relationship between landing with power and tailstrikes.

Centaurus’s description of the pitch up during go-around is a bit alarming, because despite the ground school guys talking about the second TOGA push, etc., nobody ever demonstrated that in the sim. I can recall disconnecting the A/P on the MD80 during a couple of real world go-arounds because I was pretty sure the A/P had entered the low-level loop mode. This is a something to be aware of for sure.

[RAT 5]

He showed quite a bit of flight test data indicating a much better and more precise speed control in gusty crosswinds, etc. with autothrottles on during manual flight.

That is not what I would have expected. I found on gusty automatic ILS's the A/T was a little sluggish and they gave a large thrust input which resulted in a trim change etc. etc. It was always lagging behind the a/c. On the other hand in manual flight, the inbuilt human sensors were operating ahead of the a/c and trends, and making gentle corrections much sooner with less power/pitch couple trim problems. Given that, on a gusty day, Vref usually had some add-on, I'd set a datum thrust and let the speed bounce up/down and leave N1% alone. My 'floor' was Vref not MCP speed. And quite often, if there was a smart reduction in speed you could sometimes sense it was about to be reversed and so did very little. I'm not sure the A/T was good at anticipation.

[Mansfield]

RAT 5, that is exactly what I would have expected. I can't post the powerpoint, but essentially he shows quite a bit more dynamic N1 variation with the autothrottles OFF vs. ON. In the one test condition that he cites airspeed data for, he shows an average of Vref+9 and a max of Vref+20 (with A/T on and bug set at Vref+5), compared to an average of Vref+18 with a max of Vref+31 (with A/T off and bug set at Vref+12 for wind additives).

I cannot yet speak from any experience on this ship, but like you, I have always found that I could do quite nicely with manual throttles. I don't know what my actual profiles would have looked like, but to me, running the throttles is just part of that pilot stuff. However, if nothing else, I think Bryan's argument is that the recommendation to turn A/T off when A/P is off probably can't be sustained for any good reason. Nothing in the Boeing manuals says we can't go all manual like we have always done. But the main argument that I have heard from our training folks is that with A/T off, you must add the wind additives to Vref+5, whereas with A/T on, you do not need to do this. Apparently, this is some kind of revelation...although I could swear it's been in the Boeing manuals since the beginning of time.

Now that I think about it, I think I smell the TALPA people around somewhere...:bored:

[Lancelot de boyles]

I started my B737 life in an operator that actively used A/T in 'speed off' mode on a manual approach. No issues, ever, despite coming from an utter dinosaur of a smaller jet.
I continued with another operator, a few years later, where the same practice was common place. Never any issues there, initially, and we flew plenty of challenging trips, and not a few go arounds which were non-events, using an A/T Armed/speed off setting. Then the training department started to change its culture, and along with protracted wordy briefings came the dictate that 'speed off' was not only no longer to be used, but was expressly forbidden. Why? Because Boeing said it Must Never be used. Oddly, training manuals didn't really reflect this, and after many queries from folks who'd never had any issues with 'speed off', a supporting 'memo' was supplied which quoted an extract from Boeing. The exact words were curious, and the main gist being that while Boeing 'does not recommend...', this had been interpreted as 'Boeing recommends do not...', and thus Boeing 'forbids...'

I moved on through several different types and operators for a few years, including one 737 operator where their go-around SOP had the PF reengaging the A/T, and the PM checking that it had been reengaged-

Here's the mouth music-
PF: 'go-around, flap-15, set go-around thrust' (PF sets A/T switch)

but found myself back on a B737 in Europe after 6 years, with a company that happily encouraged the use of 'speed- off', but also accepted A/T off if preferred. And so life continued for a short period of time until some folks from a Golden-harp operation took over a UK located/EI AOC, and all of a sudden, an edict was passed that clearly stated that A/T MUST be turned off, and that 'speed-off' must NEVER be used, and as further fact supporting this, an extract of a Boeing statement was issued as part of a flying notice. The old familiar words 'Boeing does not recommend...' being interpreted to state 'Boeing recommends do not...' and that the use of 'speed-off' is expressly forbidden (ostensibly, by Boeing). Along with this came an implied threat that anything else would be seen as a deliberate breach of SOP.

I'm now with another operator where this very same 'interpretation' of the by now familiar Boeing memo exists, but with a caveat that it doesn't apply if there is appropriate training in place to cover the use of 'speed-off'

I've seen plenty of notes and ops manual entries, etc, where the recommendation of completely turning off the autothrottle when flying manually is concerned, and I've seen plenty of extracts quoted regarding the 'Boeing does not recommend' or similar. But if Boeing really did forbid its use, why would there be a caveat about the training of its use? Why would the system still allow its selection (..or deselection)?
To me, it strikes me that a rigid, but flawed, interpretation of a memo addressing a 'general' technique has led to an equally rigid and somewhat flawed SOP and a degradation of the operation.

but that's just me.

[RAT 5]

Lance, I think you've come across some folks who think they know everything & don't have the time or interest to question or consider what they don't think has any necessity. Sadly they are happy to either live in the dark or reinvent the wheel.

[Mansfield]

The word "prohibit" should only appear in the Limitations section. You know, like "Use of aileron trim with the autopilot engaged is prohibited". Otherwise, I've never seen Boeing "forbid" or "prohibit" anything. This is really a social convention...in which, when you haven't got a good enough argument for why you argue some point, you default to a higher authority. Otherwise known as "appeal to authority", or argumentum ad verecundiam, a logical fallacy common enough to be discussed by St. Thomas and John Locke.

But it greatly warms my heart to see that my own training department is not alone in their ignorance of what this thing does and why...;)

[RAT 5]

Otherwise, I've never seen Boeing "forbid" or "prohibit" anything.

Indeed. I've flown only Boeing airliners, of various types and for various operators in different countries. It was interesting to see the differences in some procedures and use of this or that. When I asked how XYZ was allowed here, and not in my previous airline, the answer was, "we asked Boeing if they had any objection to this or that and they said no." So a 'not recommended' or some such became an OK if you asked properly.

If anyone can accept a little drift I have a few other technical questions regarding other 'not recommended' and wonder about what operators do.

[Lancelot de boyles]

After over a decade of hearing rigid responses regarding vague, or maybe 'broad' recommendations, a little thread drift might help clarify a few others.

[tdracer]

"we asked Boeing if they had any objection to this or that and they said no." So a 'not recommended' or some such became an OK if you asked properly.


Not quite - when I was still working we'd often get requests to do something other than the Boeing recommendations. We'd look at it, and if we couldn't find a reason to prohibit it, Boeing would provide an "NTO" - No Technical Objection. NTO is not the same as an OK - the NTO give Boeing plausible deniability if something does goes wrong (you didn't follow our recommendations, even though we didn't tell you that you had too). While the difference may not be important to the pilots, it's very important to the lawyers.

[RAT 5]

tdracer: Indeed, that is correct and I was less than accurate in my choice of words. It sounds like you may be able to answer some other ones then.

Back to Lance: On previous Boeings our 2 eng GA procedure was the same as a normal takeoff. An engine out acceleration was the same. i.e. a stressful less practiced manoeuvre was the same as the everyday well practiced low stress one. There are 4 profiles, takeoff EFATO, 2 eng GA, SE GA. They could all be the same but our B737 SOP's had differences. Takeoff has SOP to engage A/P at 1000'. Low stress if in crowded or claustrophobic airspace. 2 engine GA SOP to fly manually until flaps up. High workload for PF & PM, perhaps in unfamiliar airspace and with a WTF moment. Why not use A/P? Done it in the sim and it works fine: did it once on the line for airmanship reasons and it worked fine. The result is that MCP speed bug controls speed and not the flap lever, the profile is the same as a normal takeoff and the workload in an unusual manoeuvre is reduced. (And, if you have flown an A/P ILS and then GA the A/T stays engaged and the A/P disengages, are we supposed to disconnect A/T during acceleration to standardise?) SOP guys said it was due Boeing recommendation.
A common story in all my companies, and on prune, is the most screwed up simple manoeuvre is the all engine GA. Why make it different to a normal takeoff? If it was as similar as possible then perhaps less screw ups. (the same is true with RTO SOP's & calls v normal landing SOP's & calls; but that's a discussion for another day)
Same with engine failure. In sim I tried it, engaged A/P at 1000' accelerated like a normal takeoff (with less V/S) and concentrated on navigation and balancing the rudder. It was so easy to manage the NNC operation like that with a clear head. SOP was manual flight until Flaps up. Same with SE GA. You might be flying an A/P on A/T off ILS then GA manually. !000' A/P on if you require. Not allowed. Same reason given for rushing to retract flaps above 400' instead of a relaxed 1000'. Qeh? It's allowed, technically, but it's not a recommendation. Don't rush is the recommendation. Why? Boeing recommendation. Yet the ops manual suggested full use of automatics as required to improve overview and management of the operation.

I may be not quite correct with "Boeing recommendation", but that was the answer given. There maybe a Boeing 'not allowed', but that wasn't said. However the AFDS allows its selection, and it works fine in the sim, and if it reduces workload and improves overview and thus safety, why would it not be allowed? Some behind the buttons guidance and knowledge would be appreciated. Thanks.

[Centaurus]

However the AFDS allows its selection, and it works fine in the sim, and if it reduces workload and improves overview and thus safety, why would it not be allowed?

Horses for courses Rat 5. I have never understood the almighty rush to engage the autopilot after take off in normal ops, single engine go-around and non-normal like engine failure after V1. Perhaps it because the pilot is under-confident of his instrument flying ability; especially if the vast majority of his line flying is with the comfortable crutch of the automatics; whether company imposed or out of personal choice.
It becomes a matter of personal opinion but is the workload so intolerable that immediate automatics relief is necessary lest the crew cannot cope.

Forever etched in my mind was the CVR of the Middle Eastern captain screaming to his equally frightened first officer as their 737 rolled into a fatal IMC spiral dive soon after take off, of "ENGAGE THE AUTOPILOT" repeated in a flat panic, because he lacked the basic instrument flying competence to hand fly out of his self imposed unusual attitude.

[sonicbum]

2 engine GA SOP to fly manually until flaps up. High workload for PF & PM, perhaps in unfamiliar airspace and with a WTF moment. Why not use A/P?

My guess is that a combination of high energy and low go around altitude could inhibit the AP from entering ALT ACQ, especially if the stabiliser is not properly trimmed and the pilot is "pushing" on the wheel during the manoeuvre. The pilot could then be distracted by trying to engage the AP unsuccessfully without trimming the aircraft appropriately. Of course there must be a combination of all these adverse factors but I believe Your operator wants to consider the worst case scenario. Maybe the same considerations can be applied for the remaining points.

[Denti]

It does it quite well if you did a dual channel approach. Which was the norm according to those SOPs that i have flown with. After all, one click on the TOGA button for reduced go around thrust is quite enough, especially with a low go around altitude.

[RAT 5]

The point is I've done them all in the sim; I've done it, 2 eng GA, on the line because the circumstances warranted it. The look on the F/O's face was magic. Afterwards he asked, "are we allowed to do that?" Answer, "what did you think about? It worked didn't it? Was it a good idea or not" I'm not an advocate of transferring command to the automatics; absolutely not; (control is not the same as command), but, when it is better to do so and 'manage' the scenario with a clear head then for me it is a tool. To be prohibited to use it when it is available, either by an SOP or a 'not recommended' is the discussion i want to have. I was taught to control the a/c via tactile dexterity or automatics. Taught and well practiced being the operative words. Being forbidden to do what I choose to be the best option, and is technically possible, is a retrograde step. I think we've been here before, many times. Perhaps I should retire quietly & gracefully, but having to teach, or in fact not teach, cadets on the TR courses is becoming more depressing by the session. The S.S S & S calls, loudly, I feel.

[1+F]

Agree with all that galdian has said. We used A/T ARM too, worked perfectly, no incidents, its function understood by the crews and the benefits outweigh the so called risks. Boeing just playing the lawyer game.