B737 Single engine & use of Autopilot.

[RAT 5]

Guys. Been having a discussion with learned friends and would like to broaden the spectrum of opinions.
The topic was use of autopilot after an engine failure on take off and after a SE G/A.
Here are a few questions to construct a poll. This is not about opinions, yet, as this is just the start of the topic and I'd like to glean some data first.

Regarding SOP's: No Dual Channel Approaches.

1. Engine failure on takeoff BEFORE autopilot engagement:
a) who is encouraged to engage A/P for flap retraction acceleration?
b) who is forbidden to engage A/P before retraction is completed?
c) who has discretion?

What are people - especially b)'s- told to do if engine fails on takeoff just after A/P engaged and before flaps are retracted?

2. SE G/A:
a) who is encouraged to engage A/P for flap retraction?
b) who is forbidden to engage A/P before flap retraction is completed?
c) who has discretion?

Extra:
3. 2 Engine normal manual G/A, or after A/P disconnect.
a) who is encouraged to engage A/P for flap retraction?
b) who is forbidden to engage A/P before flap retraction is completed?
c) who has discretion?

The 'note' question applies for G/A's scenario also.

I know there are operators who specify, SOP, one of the 3 options; I'm curious at the ratios.
I've flown a) & b) SE in the sim for my own experimentation. a) works great. Engage A/P with the a/c nicely in trim, then accelerate and control a/c via HDG SEL/LNAV and fly the rudder. Very relaxed, both for PM & PF. In other words the a/c is designed, technically, to allow this. b) is OK but requires more skill, concentration & workload for the crew. If you operate under SOP b) I know there are young captains who think it is not allowed, technically, when it might save the day, one day.

FCTM: 'Autopilot Engagement-One Engine INOP'. "when at a safe height, >400' agl, the autopliot may be engaged........"

The reason for asking Extra 3 is if you operate under b), but use the A/P during a normal takeoff, why make a G/A different from a normal takeoff? Is that not making it complicated and increasing workload for the crew. 2 engine G/A a) works just like a normal takeoff; great.

There are a couple of threads about automation dependancy and lack of manual skills. Comments are that manual skills should be kept sharp and knowledge and operation of automatics should also be kept sharp. With that in mind I believe that training should emphasise both aspects of any scenario. i.e. you should be given the knowledge and training to choose which is best for the scenario you find yourself in. Recurrency training should maintain a standards check of both. Thus I'd vote for c).

I'm curious about the different SOP cultures out there. Replies can be titled 1:a,b,c. 2:a,b,c. 3: a,b,c. with simple yes/no. It would be great to hear opinions later on after we all have a better overview of what's really happening.

[parkfell]

Are not the answers dependant upon the software in the FMS. And whether classic or NG? No dual channel approach when OEI.
Fail passive would give?
1. B
2. B
3. A

I suspect the AP would cope OEI prior to flap UP.

As for fail operational...........

[FlyingStone]

There is a paragraph in 737 FCTM that recommends manual flying until the flaps are up in OEI operations in order to keep ADFS in TO/GA mode during acceleration and flaps retraction.

[RAT 5]

Flyingstone: Thanks. Indeed. FCTM says with FMC 10.7 it is normal; it doesn't say prohibited. That's my point. Plus, FMC 10.8 is now out there. The FCTM does complicate it by mentioning that with FMC 10.8 and arming VNAV for takeoff you normally engage A/P above minimum height. Complicate, because what if you have 10.8 and don't arm VNAV?
I'd always been told that TOGA has a better ratio for acceleration v climb than LVL CHG; that was all. When bugging UP in TOGA the FD gives 0-200fpm ROC. With A/P engaged, not in VNAV, will it be so very different? If you get more climb than acceleration in LVL CHG the only -ve to that is you'll take a little longer to clean up the flaps.

Parkfell: No dual channel operations in my questions.

[Jwscud]

My answers for a big EU Loco:

1. B
2. B
3. B

In practice, on real world two engine go arounds I have seen the autopilot engaged early to reduce workload as often as I've flown or seen them flown manually.

The other (biz) jet type I've flown advocates putting the AP in asap. It also had a rather basic autopilot and level acceleration was done in ALT HLD before climbing in SPD at Vfto

[RAT 5]

Jwscud. Thanks for the info:

In practice, on real world two engine go arounds I have seen the autopilot engaged early to reduce workload as often as I've flown or seen them flown manually.

Indeed I was in a 3B airline. The Wx was foul, hence a G/A. No radar, very low MAA level off, other traffic about and I wanted to execute a G/A routing different to standard due TS. The A/P went in at 400'agl and we cleaned up; went to a Hold in clear air and considered our options. Afterwards, with time, the youngish F/O expressed astonishment that the A/P worked and how easy it made the whole situation. He truly thought it was not allowed.
Spoke to a friend flying for a 3B operator who made a G/A in CIA RW33 after a circle. No radar at CIA (it is at FCO) and a very complicated G/A with low level MAA. He told me it was scary chaos as CIA was shouting to demand the reason for G/A. The crew was trying to establish exactly where ATC wanted them to go for a 2nd attempt, which frequency they should use, all the while F/O was manually flying the a/c and trying to command clean up. Ouch.

[de facto]

FCTM: 'Autopilot Engagement-One Engine INOP'. "when at a safe height, >400' agl, the autopliot may be engaged

When they write "safe height",they mean MFRH.
You need to keep TOGA thrust (SE TO(except in vnav t/o)or Simple 2 Eng GA) until you have cleared the t/o path obstacles.
If you engage at 400 feet(technically possible but if 400 is below MFRH) you will engage a pitch mode that will aim for your mcp speed rather than the required GA pitch,you need that pitch to satisfy obstacle clearance.
Your thrust,if AT is engaged will go to climb thrust rather than the required t/o thrust until MFRH.

Engaging AP below MFRH is poor practice that is creeping into practice and normally pushed by trainers or airlines who have no trust in their line pilots handling of a basic pitch with FD.
Engaging the AP which opens the mcp window will require also more work from the PM who will have now to select UP speed or required speed,rather than just retracting the flaps step by step.

Engaging as soon as obstacles are cleared,depending on the complexity of the GA maneuver,airspace... could be done,but briefed for obvious reasons.

I prefer to re engage the AT first after flaps are up then and only then re engage the AP.

So im a BBB .....unless situation dictates AP ON for easier position /maneuvring awareness(ie complex EOSID/GA) but never never below MFRH.

[Denti]

1, 2, 3: All are A in the companies i have flown for.

However, especially for case 2 and 3 dual channel is a factor, as the SOP is to use dual channel on every approach, even if it is planned to be flown manually later on (same now on the bus btw). And of course we used VNAV takeoff as soon as 10.8a made it possible. Additionally, dual channel of course uses the rudder channel servo until another roll mode is engaged which makes it the preferred method during S/E go arounds, just be ready for the kick on the pedals when the servo disconnects.

Usually i got scolded during debriefing why i didn't use the autopilot earlier on the usual V1 cut. The go around with autopilot on had to be trained extra, as the usual S/E circuit demanded that the go around and second approach had to be flown manually, usually without a FD as well.

[de facto]

With ceiling at 300 ft (cat 1 airfield)would you set up for CAT2/3?When would you disengage?
I dont understand this dual channel nonsense for every approach...thats again bean counters worry of their pilots lack of skills.
This will only lead to more when their crews rely on APs for go around and one day they disconnect for some reason,they will end up losing control.

[RAT 5]

To my understanding Boeing MFRH is >400' agl, but for training purposes they suggest 1000'.

I have not suggested engaging A/P below company MFRA, usually company minimum 1000'agl.

TOGA is a speed mode not an attitude mode. On take off that speed will depend if you have an engine failure or not, and at what speed the engine fails. The MCP window stays open. On SE engine G/A the FD is initial pitch and then MCP speed which is Vapp. The window stays open. On normal 2 engine G/A the window closes and speed bug is F15 speed for MTOW. This is a higher speed than necessary for given ALW F15. Engaging A/P with MCP wind closed, it then opens MCP window in no worse a climb gradient than you had before. By retracting flaps >400' in automatic speed mode, the a/c will reduce attitude to accelerate and reduce ROC. This may happen before you are clear of obstacles, but is a Boeing procedure and adopted by some companies. Where does the raw pitch attitude command come from?
By engaging A/P at 1000' on 2 engine G/A allows PF to bug UP the same as a normal takeoff. Then PM can retract the flaps in a calm manner as you do every day. By observation, in the sim the biggest screw up of a normal manoeuvre is 2 engine G/A, when it is done different to a normal takeoff. When done in the same manner that is not the case.

[Denti]

@defacto: we are encouraged to fly raw data approaches whenever we feel like it. On the 737 around 60% of my approaches were flown that way. However, if we do elect to use the autoflight system for even a part of the approach it is dual channel, even if we switch it off at 2700ft (usual platform altitude here is 3000+). Makes it much easier to fly a missed approach if one becomes likely as the autopilot remains engaged, unlike with a single channel approach. And flying a missed approach in places like for example ZRH is probably not the best place to show my superior flying skills but rather to show my decision making by using the available means.

[Derfred]

C, B and C.

However, for takeoff, only with update 10.8 or greater and only with VNAV armed for takeoff (LVL CHG is not an appropriate mode for S/E flap retraction, hence not used for S/E G/A).

Having said that, we've only just started practising S/E takeoff with autopilot at 400' in the sim, so I doubt many of us would bother attempting it in a real situation until we are more practised at the technique. But it certainly works well, provided particular attention is paid to trim.

I think it's great to have the option, and to practice both - there may be times when hand flying is appropriate and also times when maximum use of automation is appropriate.

For example, suppose it's not JUST an engine failure. Maybe you have a smoke filled cockpit and a face full of windshield glass from the flock of geese that also took out the engine and you look around and your colleague has been decapitated by said geese, your bifocals are smashed on the cockpit floor along with your headset, and the fire warning is still blaring - there is a little button that can make your life a little easier: A/P CMD).

[RAT 5]

LVL CHG is not an appropriate mode for S/E flap retraction, hence not used for S/E G/A.

I hear this from various people but can not find chapter & verse reference to it. Can you please provide Boeing reference to explain why.

[Jwscud]

I can't find a Boeing reference, but I would guess it is because LVL CHG will not command a level acceleration segment but up to a 500fpm climb at least (watching acceleration two engines when cleaning up &c.)

Denti - how does your company permanent dual channel policy deal with disconnections below 400ft on landing? You are then manually flying an aircraft that's bloody miles out of trim which is hardly conducive to a stable approach.

I have only flown asymmetric for real once (in a bizjet due to an ECU failure) and I was far more comfortable flying manually once the QRH nonsense was complete due to being more in the loop and the nature of the failure. An autopilot without a rudder channel makes for a curious mix of automatic flight but manual handling of yaw and rudder trim and my view is when thrust and speed changes are involved it is wiser to fly manually than risk the autopilot kick out with a large handful of out of trim forces.

[RAT 5]

I can't find a Boeing reference, but I would guess it is because LVL CHG will not command a level acceleration segment but up to a 500fpm climb at least (watching acceleration two engines when cleaning up &c.)

But is this correct or rumour? Is it a problem? I notice, in the sim, that with an assumed temp takeoff, where minimum thrust is used for the weight, that for SE acceleration at MFRA the FD (TOGA) gives level flight. Now use greater than necessary thrust and you see a slight climb depending on the surplus thrust. I was always told (true or not?) that 0-200fpm was acceptable in the calculation.

Denti, for an EU operator, is allowed to use autopilot for SE flap retraction. Therefore it is approved by XAA's. It is allowed by Boeing. All I'm trying to understand is why it is taught by some that it is forbidden when that seems not to be correct.
This links to the topic about lack of knowledge of a/c systems especially in AB's and confusion amongst pilots. Many believe their SOP is gospel; but now we learn other colleagues are taught differently and use the systems within their design parameters. IMHO there is too much blinkered training and not enough general training. Is it healthy that crews are taught only a very restrictive method of operating their a/c? Why are they not taught the design parameters and then SOP's?

[Derfred]

I hear this from various people but can not find chapter & verse reference to it. Can you please provide Boeing reference to explain why.

FCTM:

For airplanes with FMC U10.7 or earlier, autopilot engagement is normally delayed until the flaps are up and LVL CHG is selected. This allows the AFDS to remain in the TO/GA mode during flap retraction.

I can't tell you if those are Boeings words or words of my operator. I note that it is not a prohibition but a recommendation. Recommendations are generally regarded as SOPs by my operator, unless an overriding safety consideration is applicable.

I believe the reason is that TO/GA "knows" you have an engine failure, and will pitch appropriately. LVL CHG does not know you have an engine failure, and will only pitch for bug speed, including possibly commanding a descent if you are below bug speed. The comment above regarding 500fpm is not correct.

[Denti]

Denti - how does your company permanent dual channel policy deal with disconnections below 400ft on landing? You are then manually flying an aircraft that's bloody miles out of trim which is hardly conducive to a stable approach.

Yup, due to this particular quirk of the 737 AFS it is out of trim. Dunno how they do it today, back when i got my typerating we had training on this. Including quite a few approaches in the SIM where we had to disconnect after the back trim was in and fly it to a stable landing. At least another one of those was later on done during line training. Although, my outfit allows FO CAT I autolands, which can help prevent the need for low level disconnection during CAT I conditions. Past 200ft (alert height) one needs a bloody huge problem to trip off the autopilot, in which case a go around might be a better option if still possible.

About the OEI case, this is what the manual (custom FCOM) in my company says:

Autopilot Engagement - One Engine Inoperative

When at a safe altitude above 400ft AGL with correct rudder pedal or trim input, the autopilot may be engaged. When VNAV is armed for takeoff, the autopilot is normally engaged anytime the airplane is above the minimum altitude for autopilot engagement.

In VNAV the AFS will usually command a shallow climb during flap retraction, but then, TOGA does, depending on thrust to weight ratio, command a 0-200 fpm climb as well. The only mentioned pitch modes for flap retraction in the S/E case are TOGA and VNAV with the latter one the preferred mode as it automatically uses the correct S/E acceleration height and has a LE protection speed feature.

[RAT 5]

Derfred: Those words are from my 2014 FCTM. I'm not sure about "TOGA knowing you have an engine failure"; certainly not before you've selected fuel to cutoff. Thus the comment about LVL CHG NOT knowing you have an engine failure may also be unconfirmed. The chance that LVL CHG might command a descent after selecting Bug UP is very unlikely.
If VNAV is an acceptable mode, engine out acceleration, is it because it is so clever. This use of VNAV on departure in 2015 is very very similar to B767 in 1995. Wow; what took the young son so long to catch up with Dad?

[Jwscud]

Unfortunately, I have a cheap FCOM that doesn't give much detail on the modes, and we don't allow arming of VNAV on the ground. TOGA clearly detects engine failure given the 3 options on commanded speed in an engine failure. It is entirely opaque to me from the documentation I have how it does that though!

The latest FCTM I have is Rev 14 (June 2015)

For the 3rd sector, the aircraft is required to accelerate at a given rate to meet the requirement to be clean at Vfto within the MCT limit time. The NTOFP for the 3rd segment is an acceleration gradient to meet that, which is what the 0-200fpm climb limit in the FD is there to command (see FCTM 3.39) - it suggests in raw data to maintain "approximately level flight"

This seems to me a bit academic given a 10 min as opposed to 5 min MCT limit is available as a service bulletin with I believe no technical changes.

[RAT 5]

OK: for those who do not use LVL CHG to retract flaps due to perceived performance problems etc.
Some say you have to use TOGA to maintain obstacle clearance; some say it is to give level acceleration at MFRH to reach Bug UP ASAP and not infringe 5 mins takeoff thrust. Surely that would only be of consequence if at Max Takeoff Thrust not at an assumed temp?
Now: you engage A/P and have a bird strike/engine failure/loss of thrust before MFRH. What do you do with the autopilot and pitch modes?
From a performance point of view you are well above the considered V1 cut climb profile, so all is OK except for the 5mins.
What do you do with a fire? Again you are well above the SE climb profile to MFRA. It is likely you'll shut down the engine and lose thrust before you start the acceleration phase.
We are seeing again that there are differences in operator procedures, even if only slightly. They are all approved by XAA & manufacturer. Back to knowledge of a/c. Many think there is only one way to skin a cat and everything else is wrong. This stems from edited training and slapped wrist every 6 months if you deviate. So, just how educated are todays pilots? Are we seeing repercussions of this dilution in some accidents?