For anyone who reads Richard Collins articles in the American "Flying" magazine (one of my mentors) He recently stated that the FAA who hitherto had always advised immediate disconnection of the autopilot in the event of loss of control in IMC, have now changed their official viewpoint to state that where an autopilot is fitted it should be the first line of defence, and promptly switched on...not off in such circumstances

Some very good points here.

In my last company we had a mixed fleet, some a/c had autopilots fitted but most didn't (or were U/S).
When you found yourself with a machine with a serviceable one, it was a cause for celebration. Hand flying isn't difficult, but for long straight and level sectors it became very tiring and boring. Having a good autopilot for us helped free up the non-handling pilot whilst the PF could monitor what the box of tricks was up to.

I personally don't like flying single crew I/R without a serviceable "George." If you understand it's limitations and how to use it properly, then they can be life savers, especially when the workload gets high. Your situational awareness is always helped by its use, since your brain has some extra capacity (though if you are experienced, the difference can be negligable compared to someone starting out).

I always like to start handflying again before the approach phase, so that I could get myself "back into it" before the most critical phase, though we obviously sometimes used it for the approach aswell, though generally not when conditions were bumpy as most old or basic autopilots tend to give up just when you need them!

If you find yourself starting to lose it in IMC, an AP can be a great help, just make sure you know when it will engage and what makes it trip out and where the disconnect button is incase it does something unexpected. (Usually a pilot problem rather than an AP problem though!)

First of all, I would guess that asking an airline pilots what he does at work, as F3G suggested, might not be the best idea

That's not what I suggested, please read my post again.

Many airline pilots are familiar with simple autopilots from their light aircraft experience and are in a good position to highlight the pros and cons.

Until one is proficient in using the autopilot it actually increases the workload.

Best to get some practice in with an instructor before using it for real in IMC.

As for not using a simple autopilot in order to retain flying skills. I would not worry. For a simple 1 axis autopilot, the pilot still has to manage the other 2 axis as well as the navigation of the aircraft. Any proficient pilot will be able to trim out an aircraft with no autopilot and fly hands off for long periods in IMC be it climbing, descending or straight and level.

It's not about flexing muscles to control the aircraft. IMC flying is all about management of the siutation and keeping one's mind well ahead of the aircraft.

If you are dealing with a full 3 axix autopilot with altitude capture, vertical speed modes and coupled navigation and approach facilities then I would agree that long periods without some hand flying would reduce motor skills. However, the mental skills which account for a big percentage of IMC operations would be stil up to speed provided that the autopilot was not being used as a crutch to support a person who lacks the basic skills to start with.

Pilots who use RANT on a regular basis will have better situational awareness regarding procedural flying than many others who do lots of aircraft flying. The reason is that RANT allows more brain time to think about what is going on. With practice many issues become almost automatic and that releases more thinking time for other things like flying the aircraft. An example of a device that is not even in the aircraft that regular use helps not hinders IMC flying.

To sumaraise; Get training from an instructor and with thorough knowledge of the system and it's legal as well as operational limitations make the most of it.

Regards,

DFC

A few points:

GA autopilots do not have redundancy, and most have poor error detection, and they can and do fail. I've had many autopilot failures, and this is the most modern Honeywell KFC225 type. So while an AP is a great safety aid (through workload reduction) one does need to keep an eye on it, and know where to disconnect it. The red button is not enough: I've had one failure where the red button didn't disconnect, and one needs to know where the AP master switch is, and/or the servo circuit breakers.

If one has a flight director then one can use the AP for that even if the servos are burnt out.

Most PPL-level instructors are clueless about autopilots and most other forms of "advanced" avionics. One is just as likely to be told a load of bull. So one needs to choose the instructor carefully. All will speak with confidence, but only some know this highly type specific stuff.

Whether an AP should be engaged upon loss of control in IMC is debatable. I'd like to see a reference for the FAA guidance reported - it must be type specific. An instrument pilot must know how to do (partial panel and full panel) unusual attitude recovery in IMC. If one's vac pump goes, and you have a vac horizon, and the AP happens to work off that, then engaging the AP will do no good. You have to recover manually, and will be doing it partial panel too. That's why one must understand the particular installation. Also any modern AP will auto disconnect of bank or pitch values exceed X degrees, say 20 or 30, and will thus disconnect in serious turbulence.

One should use an AP all the time conditions of actual or potential heavy workload. Always engage it if you have too much to do. Then you won't get into a loss of control situation in the first place.

All good stuff IO540. Perhaps a few ILAAFFT autopilot incidents might not go amiss.

1. After several years of using a fully coupled attitude based 3 axis A/P with altitude hold and capture ( separate Yaw damper) and automatic trim, I converted to a rate based 2 axis unit, again with altitude hold but no auto trim. In the cruise with the A/P having been engaged for some time, I decided it was time to descend, only for some obscure reason ( I think because the trim wheel was not periodically moving and thus giving me a constant reminder that the A/P was engaged) I forgot that it was indeed actually engaged and reduced power for the descent. I was surprised when instead of the nose dropping, it remained pegged to the horixon (altitude hold engaged) I therefore attempted to move the stick forward and found to my horror it was apparently "Locked." This was followed by winding in a fair dose of forward trim and a further power reduction in order to encourage a descent. Still the nose remained pegged on the horizon and airspeed began to fall rapidly. Realisation suddenly dawned and (ignoring the trim warning lights) I instinctively switched off the A/P resulting in an immediate severe bunt nose down which pitched myself and the front seat passenger up to the ceiling. Fortunately the event occurred in good VMC and immediate manual recovery was instinctive and uneventful. One only needs to imagine a reverse scenario however, where an attempt to climb preceded by a power increase, a tug on the stick, a large dose of aft trim and immediate A/p disengagement would result in a pitch up stall spin scenario which in IMC would have obvious repercussions. Lesson learned, Know your A/P inside out, read the operating manual, understand what mode it is in and avoid the "whats it doing now" situation.

2. Flying R/H seat with another (experienced) pilot in the left hand seat we reached 200 feet on the climb out when he announced that "Strange forces" were acting on the ailerons. I took the stick and immediately realised we had taken off with the A/P engaged. After a quick check of the engagement mode I disconnected and the flight continued uneventfully. The implications were obvious however. a) An unsatisfactory pre flight check had been carried out with failure to notice the A/P already engaged. b) Had the previous pilot actually therefore landed with the A/P engaged? (Far from unknown-imagine battling a cross wind landing against an engaged A/P and yaw damper) c) If altitude hold had been engaged, an attempted takeoff with the pilot commanding a climb against an A/P resolutely determined to maintain runway elevation poses an interesting end of runway situation.

3. Most A/P's have a test or self test facilty which together with an operational check must be carried out prior to departure. First time engagement of the A/P in the air without a ground test first is not a good idea and can result in some A/P's inducing some very dramatic attitude excursions in the air!

As others have said it is not usually the A/p which is the problem, It's the pilot!

True indeed and most GA autopilots are relatively easy to overpower. This again is checked as per the operating manual eg when the autopilot is set to command a left roll, the pilot ensures that a right roll can be executed against the autopilot without a need for excessive force. Generaly autopilots that control the elevator electric trim system trim are set so that the trim circuit breaker will pop with a runaway trim. This should happen in the case of at least one system when the out of trim forces are about 15Lb. Not a great force to overcome but one which as said above could catch one unawares if the autopilot was disconnected while not holding on to the controls and not properly strapped in.

Reminds me of the story about the pimply computer whizz first officer and the crusty old Captain in the new Airbus some years back. First officer operates the sector totally by jabbing away at the flight management system until at top of descent the aircraft does not descend! Much trashing of keys and a shocked statement of "It's not descending Captain" resulted in the Captain leaning forward and pushing the yoke towards the pannel. "It's descending now son isn't it!" :D

Regards,

DFC

I have been a regular autopilot user in both VMC and IMC conditions, the key is to continue to pay attention to what's going on. You are still the pilot not a passenger being flown by a piece of kit:eek:

I've flown both altitude hold and none and my current aircraft has essentially the wing leveller.

My wing leveller can also follow the heading bug and do VOR tracking and approach interception unfortunately it's got the dreaded wing rock at present so it's not useable in IMC and there's no doubt that keeping up without it is bl**dy hard work in IMC:mad:

A little off subject perhaps but, my aircraft is equipped with a 3 axis A/P. When I first bought it and was doing ‘differences’ training, my instructor covered the use of the A/P. As I am not IR rated, he told me “If you inadvertently enter IMC turn this thing on, use it to do a 180, and get back to clear air”.

I just read an article about inadvertent entry into IMC. It said they tested 10 none IMC rated pilots suddenly entering IMC conditions. The average time to loss of control was 170 seconds! None lasted longer than 8 minutes. They advised to always properly trim the aircraft so it will fly S & L ‘hands off’, if you do inadvertently enter IMC, let go of the control column, and if you have an A/P engage it and use it to get back to clear air. If you don’t have an A/P use the rudder only to turn the aircraft. The article also stated 40% of fatal accidents involved loss of situational awareness as the prime reason, or a contributing reason to the accident.
Seems to me it’s a piece of equipment that, properly managed, can save your life as well make it easier on those long flights.
Regards,
W.B.

Exactly what I meant - but I expressed it badly!

White Bear's comment seems v. well worth while to convey to non-instrument qualified pilots? Many will fly aircraft with some sort of A/P.

This is digressing a bit, but I too have read the various "studies" (mostly in the USA) about how long non-instrument pilots take to die in IMC. I read with incredulity about how fast pilots with 2000 hours TT lose control the moment they enter IMC. They must have spent their entire life flying circuits in Arizona! As someone who started instrument training very early on, I just don't get this; getting enough training to be able to fly a heading in IMC is no rocket science; you can teach someone to do it adequately in an hour or two. Once, I had a passenger friend who never flew a plane before, doing that much after about 10 mins' training. The other 98% of instrument flight training is all the procedures, doing a heavy radio workload at the same time, partial panel, etc. but you don't need to know that stuff to stop yourself getting killed. Unfortunately, PPL training does just the 180 in IMC and everywhere we are told to be !!!!! scared. I think this is counterproductive, but I suppose that the moment they expanded IMC training within the PPL some traditionalist old fart will stand up and say the stuff should not be taught unless it is taught "properly"....

Back to autopilots, it's an unfortunate fact that post people reading this will have only ever seen ones that (at best) are placarded INOP and (at worst) are duff and go out of control as soon as you switch them on. They are expensive bits of kit and the first thing most renting establishments do when they get a problem is to pull the circuit breaker and forget about it. "The plane is for VFR training after all". In my 2 years of training/renting I saw a number of autopilots and not one of them worked. Keeping a plane with decent avionics, all working, is a serious financial commitment.

Having trained alot of non instrument rated pilots for IMC's and IR's, the comments about people losing control rapidly are unfortunately borne out by my experience.

The very limited amount of training given in a PPL syllabus in no way adequately equips someone for real and inadvertant flight into IMC conditions. Generally most people do tend to "lose it" in a relatively short time, even when sitting next to me as a safety net. Now imagine it for real, with all the stresses involved and I can well imagine loss of control happening pretty sharpishly for the "average PPL".

IO, remember there are very few PPL's in Europe with even 1/10th of the 2000hrs you quote.

If you want proof of losing it in IMC, just look at what happened to JFK jr.

I'm generally in agreement, but what do you do when autopilot engaged and the ride gets rough in IMC. Do you disengage and handfly? Always seems to occur in the descent phase when workload is on the increase.

The normal procedure for flight in moderate or worse turbulence is

1) set speed to Va or less

2) maintain wings level / maintain heading

3) maintain angle of attack (or pitch attitude if you like) and accept any altitude fluctuations

One can do the above with most modern autopilots (provided it has not disconnected as a result of the turbulence); what one should not do is fly in altitude hold mode because that will create extra stress on the airframe.

Same procedure if flying without an autopilot.

SAS, I don't disagree with you. I just think that things could be improved. As regards JFK JR, doesn't that lead to an argument that there should be no night flight without an instrument capability? I doubt he knowingly embarked on a flight into known IMC, but it's easily done at night.