The case for regular manual reversion training on the 737.

[A37575]

During type rating training in the 737 simulator, the Boeing syllabus dictates just one manual reversion landing, and that is probably during the session that introduces hydraulic failures in turn leading to manual reversion. In cyclic or recurrent training it is rare to revisit manual reversion.

Engine failures/fire warnings at V1 are regular simulator training items and certainly there is no shortage of one engine inoperative landings, go arounds, VOR/ILS, circling and so on. And should they happen in real life, it is hoped most pilots would cope successfully.

A manual reversion ILS approach and landing in the 737 requires considerable flying skill. One could argue it is the ultimate in old fashioned stick and rudder skills - and we all know how stick and rudder skills have been degraded nowadays by the accent on automatics.

While the majority of competent pilots would be confident of conducting a safe single engine landing during a line flight, it could be a different story should one be faced with a real manual reversion landing. If a single engine approach becomes unstable, a go-around is no sweat to a competent pilot. But I wonder how many would be confident of conducting in real life, a well executed manual reversion landing. Methinks not too many...

The manual reversion approach and landing is arguably the riskiest and most prone to go wrong manoeuvre a pilot will experience in his career; no matter how remote the possibility. During manual reversion simulator training, one observes more heavy landings following unstable approaches than any other manoeuvre in the Boeing syllabus.

If this is a common observation, and it probably is, then surely competence at manual reversion landings should be practiced during recurrent and proficiency checks in the simulator? Not just once, but several times each year.

After all, which manoeuvre is the more critical and difficult maneuvre to fly? One engine inoperative landing-or manual reversion landing? Most pilots would opt for the latter even though both are only remote possibilities

[Rainboe]

I've flown the 737 for 10 years altogether. I was shocked at how difficult manual reversion actually is. Nothing to do with stick and rudder skills, it feels like nothing like it. This is an emergency 'try and keep flying' procedure. Control is VERY difficult, I think impossible to do any more than simply keep flying. Even a very experienced 737 pilot would be hard pressed to actually even hit the runway, let alone 'land' IMO. I think it is effectively only an airborne process to keep flying. As for landing, I do not think you should expect to remain in one piece and walk away from it.

It remains an extremely unlikely probability. Many twins have no manual reversion at all, so at least the 737 gives you a chance....but no more than that! It's impossible to explain what it feels like to a non pilot. I just think your odds of survival are fairly low when you come to try and place it on the ground.

[kenparry]

I endorse Rainboe's comments. I flew the 737-200 for 13 years, and never heard of a manual reversion due to tech problems in all that time. Once, I went into manual by choice - FL100, 250kt IAS, for a trim check prior to deep maintenance. It was nothing like the sim, very much heavier, so heavy in roll that with full strength I got a roll rate of about 5 deg per second. Pitch was not quite so bad. As for landing, I would try to find somewhere calm & CAVOK with a long runway. I can't see any training value in spending time on something that is so unlikely to happen. If you have spare sim time, try an engine failure just after lift-off - it's harder to handle than a cut at V1.

[rubik101]

While I agree that the liklihood of finding yourself in Manual Reversion is very remote, I do not agree that your flying career is over. I find the tone of the prevcious posts somewhat over-dramatic and alarmist.

Use the QRH.
Find a long runway, use Thrust and Stab trim for pitch control, run the Standby rudder to provide gentle yaw inputs, causing roll.
Plan a Flaps 15 landing.
Fly the ILS from 15 miles out and around 4000' Get the Gear down early, before the 15 mile point. Set up a normal rate of decent and keep it on the centreline with rudder inputs, very gently.
Don't flare, simply remove all the power at 50', let the speed reduce and allow the aircraft to sink on the runway.
Whatever you do, don't flare!

It really is not so much of a drama, having tried it several times in the last 29 years on Boeing 737s.
Doom and gloom and certain death it ain't!

[Rainboe]

I have a strong suspicion the simulator is the simulator, and real life will cause a considerable surprise. It would be interesting to hear if manual reversion has been sampled in development and maintenance flying with non-Boeing staff pilots, and how similar it is felt to simulation. The impression I get is the aeroplane deviates in roll and pitch, and it is impossible to fly steady state. I think an approach would be difficult. True, it's better than nothing (like the 757), but I always felt I could get it near a runway, but at what ROD and position would be down to luck only. I believe it has never been done in anger?

[Denti]

We do manual reversion flight training at least once every three years. We are required to train all major system failures in a three year rotating system and manual reversion is one of them.

However real life is very much different from the simulator in this case since the simulator is unable to produce the heavy forces required to fly the real thing. It is a normal check item on post D-Check acceptance flights or acceptance flights when an airplane is returned to the leasing company. Quite interesting to do those flights actually, provides a real learning experience.

[kijangnim]

Greetings
One pilot handling the pitch, the other handling the roll and avoid thrust changes if possible :ok:

[TopBunk]

Did a post maintenance test flight on a 737-200 about 15 years ago that required a test of manual reversion.

Very interesting, quite like the sim model from recollection ... big null area in pitch, very heavy in both roll and pitch, found it impossible to fly level turns and descended at about 700fpm minimum with turns. As I recall, I did not use the standby rudder to assist in the turns though - that may have helped.

With the level of training given routinely I concur with Rainboe, I would look for the longest, straightest runway in CAVOK conditions with no crosswind and pray.

Not being sexist, but I doubt if some of our less strong colleagues would be strong enough.

[Tee Emm]

One pilot handling the pitch, the other handling the roll and avoid thrust changes if possible.I tried to wrap around the above paragraph but the icon did not operate.The question of a two pilot effort sharing the task of handling a manual reversion was the subject of a letter to Boeing recently. The reply was that there is no Boeing procedure that requires the use of two pilots to fly the aircraft. In fact having two pilots sharing thrust and airspeed control is a recipe for an almighty balls-up. Certainly the FCTM advice is good about considering use of stab trim to aid the flare. During my own type rating with a Boeing flight instructor from Seattle many years ago, we conducted an inflight manual reversion circuit and final approach to 500 feet. At 500 feet the hydraulics were re-instated for obvious reasons. The feel of the aircraft was identical to the practice in the simulator a few days earlier. Naturally it took careful flying but as the original poster said, if you were competent at normal hand flying then the manual reversion was well within the skill range of a competent pilot. A Pan call landing perhaps, but certainly not a Mayday situation unless you seek and glory in the media publicity and like to walk down the stairs to a hero's reception, garlands and dancing girls etc. I agree the manual reversion should receive regular attention in the simulator especially if you feel you cannot cope with it safely in real life. The fact the event is a remote possibility means nothing. Dead stick landings are remote possibilities in jet transports but they have happened.

[Rick Studder]

Sure, manual reversion would be a very serious situation, but so would loosing all engine power and there's not much training for that either. When you assess risk, apart from the severity of the situation, you at least have to take into account the probability of it occuring.

In that perspective the argument can be made that more sim training should be used for the mundane situations, such as ordinary go-arounds. I actually think this latter focus would prevent more accidents.

[Rainboe]

If I ended up on manual reversion in a 737, I would be yelling 'Mayday' so many times I would be hoarse! An uncertain reconnection with terra firma, not even sure you will still have the plane with undercarriage still attached, let alone in one piece, and you think a 'Pan!' is adequate? Not me!

[rubik101]

Rainboe, you seem to be under the mistaken impression that the aircraft is almost unflyable, when that is most definitely not the case. In the past, with other airlines, conversion training for all pilots included training in manual reversion flying. To the best of my recollection, none of them ever made anything other than a very survivable landing. The touchdown is not any harder than a normal landing, and I find it hard to understand why you should think otherwise.
The aircraft is flyable using a combination of the flight controls, standby rudder, elevator trim and thrust, all of which function adequately.
Extra time and planning for early gear extension are required but other than a longish runway, a normal ILS or visual approach will suffice.
Do not try to make it into an almost unservivable drama when it is evidently not the case.

[Rainboe]

You're talking simulator training. I am glad you have full confidence in a successful outcome. Just flyable it may be, but I do not feel confident about the touchdown bit. The idea is not to be fluttering about wearing wings and whinging 'but it's not like that in the simulator!'.

[Stan Woolley]

I've done it a few times in different sims and the results were not consistent.

I remember on the 737-700 seeing it crashed twice on landing and thinking that it was an extremely serious position to be in. Quite recently I did it in a 737-800 sim and it was no real problem at all.

Simulators probably vary quite a bit in setup even though it shouldn't be that way. I guess the more important feedback would be from pilots who do it in the a/c every now and then?

Mayday all day long IMO.

[CaptainSandL]

Recurrent Training

I think everybody should get manual reversion training in the sim during their conversion so that they know what to expect if it should ever happen to them. But because the probability of occurrence in real life is so small it cannot be justified as a recurrent item. If sim time was unlimited then OK, but alas it is not.


Sim Fidelity

I would estimate that I have got at least 20 hrs of actual 737 Classic & NG manual reversion time. All of which was in the controlled conditions of airtesting; ie Classics at FL350/M0.74, NGs at FL150/250kts. I have probably got less than 2 hours of manual reversion time in the sim. In the real aircraft I have done some demanding manoeuvres but mainly just primary control response checks and certainly nothing at low speed or with flaps extended.

My experience of sims in out-of-the-ordinary configurations is that they give you a good flavour of what the aircraft is like, but that they are not exactly the same. Notable examples of this are taxying, first 50ft of take-off and landing, none of which are quite the same in the sim as the real aircraft. The modelling of these features can also vary from sim to sim. The same is true of manual reversion in a sim, it is close enough but not exactly right.


Manual Reversion Tips

I wrote a long piece on this before on these pages, but to recap:
• You will probably need to use both hands on the yoke because the elevators & ailerons will be very heavy.
• The rudder should still be powered so use rudder (gently) to assist your roll into & out of any turns. Note that L rudder may be reduced depending upon your RSEP status
• Just like s/e flying, try to make small changes of hdg/spd/alt early, rather than having to make large ones late.
• I don’t endorse the earlier suggestion of one pilot for pitch and the other for roll. IMHO one pilot handling and the other monitoring, otherwise nobody is monitoring.
• If you are getting tired (it is very tiring), give control to the other pilot for a few minutes to take a breather. This will also give him some handling experience which you may have to call upon for the landing.
• When you are on datums it is very stable so you should try and relax your grip (& unclench your teeth!) between manouvres.

IMHO it is survivable, most people manage it in the sim.

S&L

[Centaurus]

But because the probability of occurrence in real life is so small it cannot be justified as a recurrent item. If sim time was unlimited then OK, but alas it is not

Excellent reply. The "probability" factor is a good point. I recall the comment made by the 767 captain of the Gimli Glider, as it was called. Readers may recall he did a good job of a dead stick landing after running out of fuel. He said that if only Air Canada (his airline) had allowed even one practice dead stick landing, he would have approached his own event with more confidence. As it was it was always going to be a hit or miss affair.

Of course since that episode, there have been a few more dead stick landings including an A340 and a Garuda 737. The probability of their occurrence in real life means they could not be justified as a recurrent item in the simulator....? Well, hello!

Some argue that there is significant wasted time in simulators with much depending on the organisational skill of the instructor. Instructors who freeze the simulator for lengthy and sometimes verbose discussions, waste valuable time. Those that permit excessive track miles on automatic pilot waste valubale time. The simulator is a training device not a comfortable arm chair ride while ficticious scenarios are permitted to be played out to a ridiculous extent - some of which include the captain discussing events with the ficticious cabin staff aka the instructor.

In terms of time management, the result is a lot of superfluous chatting between agencies while the minutes tick over before the morning tea break. Much of the briefing and chats could just as effectively conducted in a briefing room and the simulator time kept for vital hands on training; ensuring competence at dead stick glide approach landings comes to mind. Include manual reversion approach and landings among that.

During one entire type rating course in the simulator not one landing (apart from a manual reversion) was conducted without the help of the automatic brakes. In other words graduates were going on to line training in the real aircraft without ever used manual braking on landing. Yet another airline syllabus of simulator training thought it unnecessary to ensure graduates had proved competent at GPWS pull up manoeuvres. Sure, they had been briefed - but never actually carried out any in the simulator. But no shortage of L-nav and V nav automatics every session.

Time and again, one reads accident reports where investigators recommend operators to include specific simulator training to prevent further events common to the accident. Occasionally an operator will take on board those recommendations - but not too often. Flight International recently quoted one European Agency as reccommending operators accent training on strong crosswind landing training in simulators following a couple of accidents involving low hour pilots flying transport jets.

One response was that the simulator sessions were already packed with essential material and there was no time for extras not considered "vital" to flight safety. Talk about a negative approach to flight safety.

There needs to be greater emphasis on giving pilots hands on skills in the simulator and this includes practice at the "remote" possibility events. Wasting valuable simulator time L-navving around a holding pattern on autopilot reading lengthy checklists does not equip pilots with the basic flying skills long eroded because of the accent on automatics. This applies especially to the new breed of Multi crew licence graduates.