The KAP140 should not do this. @Jesse might have ideas… sure do start another thread.
The King autopilots can be very good (my KFC225 works great in turbulence, holding altitude to maybe 20ft in chop bad enough to make it hard to twiddle the knobs) but they will disconnect when pitch OR roll exceeds certain values. That was a certification requirement. They differ but are in the region of 20 degrees for pitch and 30 degrees for roll.
In IMC and in turbulence one really needs to keep an eye on the AI and everything else. Still much easier than hand flying.
Hello!
PetitCessnaVoyageur wrote:
What strikes me, when considering this evidence about AP malfunction, is that we never see his hands on the controls.
That topic would be a discussion on it’s own…
First to the pilot in this video: He really can’t decide what to do with his left hand. If you look at the last five minutes of the approach, half of the time he has the hand on the stick, sometimes close to it and sometimes he lets go completely. A clear indication of how nervous this situation made him.
Personally, I am strictly against having a hand – one hand only in any case! – on the yoke while the autopilot is flying. With the one exception of aircraft which are known to be out of trim when the AP disengages. Either because the autopilot is so primitive that it doesn’t operate the trim in the first place or because it tends to be defective, like those found on the typical training aircraft…
But with a properly working autopilot, what benefit can come from holding the yoke while it is controlling the aircraft? Turbulence is a random process. After every upward jolt, a downward jolt will follow, cancelling the previous one. On average, the aircraft will fly along a straight line. This is why you don’t “steer” your aircraft through turbulence when flying manually. You hold your controls more or less steady and only correct a trend which you can see evolve, e.g. loss of altitude due to a downdraft. Letting go the controls briefly will have almost no effect. The autopilot is doing exactly the same. There are filters built into the circuits (either by analog circuitry or by digital programming) which are applied to the input data (attitude, altitude, airspeed, …) so that all the short-term variations are suppressed. Only the general trend will be acted upon. So if your AP disconnects due to turbulence, exactly nothing will happen. You have plenty of time to get your hand on the yoke and take over manually.
On the other hand, having your hand on the yoke when the autopilot is flying can induce a constant small force in the control circuit (unless you have an armrest which lets you hold your yoke completely free of any force). This little force will be trimmed out by the AP, but the moment you disconnect it, it will reappear as an opposite out-of-trim force. Who doesn’t believe it, just try it out on your next flight! I have never flown an Airbus, but I am told that the fly-by-wire system senses any control input on the sidestick and disconnects the autopilot as a result, so having a hand on the stick while the AP is flying is not good in these aircraft.
The worst case of autopilot disengagement is an out-of-trim condition which exceeds the force of the servos. That would be 20 force pounds in the aircraft I am flying. Ours gives plenty of warning if this is about to happen (“AP pitch mistrim” or “AP roll mistrim” warning lights with flashing Master Caution light), but imagine a sudden AP disconnect like this. Your hand, which is ideally placed loosely on the yoke, will instantly feel those 20lb, which equals a 10kg shopping bag suddenly dropped onto your thumb or forefinger depending which way the yoke moves. Only Superman will be able to hold that, with everybody else, the yoke will immediately slip out of the hand, possibly even dislodging a joint in thumb or finger in the process. In the end the same result as if you would have kept your hand away from the yoke, only with an additional chance of getting hurt…
Our company SOPs call for a hand “close to” the controls when flying below 2000ft AGL. But apart from that, either the AP flies or the pilot, never both of them.
what_next wrote:
But apart from that, either the AP flies or the pilot, never both of them.
Thanks for this advice. I used to put hand on yoke when flying through turbulence (because I was taught so) but after reading your post I’m sure I’ll never do it again.
@what_next : interesting insights, thank you.
Will an autopilot apply a full control deflection? Presumably this is why people are refering to exceeding Va rather than turbulance penetration speed?
I’d never be bothered about flying above Va in turbulence, but wouldn’t fly above turbulence penetration speed (yellow arc). But I never use the autopilot (other than occasionally to check it still works!)
dublinpilot wrote:
Will an autopilot apply a full control deflection?
Yes. If it can be achieved within it’s force constraints (those 20 force pounds for the Part 25 aircraft for example). Certification requirements should however ensure that above Va the control forces required for achieving full deflection are greater than what the autopilot servos will be able to produce.
If you want to see your AP apply full deflection try this little experiment: Climb to a safe altitude, engage heading and altitude hold mode and pull back the throttle. Most autopilots will be stupid enough to pull the yoke fully back in order to maintain the altitude. Close to the stall, control forces are very low so it can easily do that.
As I wrote above, the autopilot does not really feel the turbulence because all high-frequency sensor inputs get filtered out. So it will never fight the chops with large and potentially danegrous control inputs.
The FAA uses the term TAA for Technically Advanced Aircraft. Flying such a thing is different than flying a “regular” one. With the regular aircraft the pilot uses the controls and gets information from the gauges or displays. With a TAA the pilot can do that as well but most of the time the pilot interacts with the system. The trick is to know when to switch from system interaction to using the controls so to say.
I find nothing wrong with the hands in the video. My own flying probably looks very similar and just as jet pilots do I mostly rest my hands half folded in my lap :-)
And bear in mind the PTT is on the yoke…
AnthonyQ wrote:
And bear in mind the PTT is on the yoke…
Yes, but one really doesn’t need that all the time. BTW: To my knowledge, Embraer is the first manufacturer of light jets who installs a second PTT switch somewhere in the glareshield so that the pilot doesn’t need to touch the yoke while the autopilot (or the second pilot) is flying. I wonder why it took 50 years until someone came up with that…
Other than that I think in moderate turbulence – and this wasn’t really bad, there were no major bank or pitch excursions, light-to moderate I would say – the AP is best left off – I think good IFR flying is ALL about using the autopilot appropriately.
Saying that “you are using an AP and aren’t manipulating the controls so you are not really flying” is equivalent to saying “you are using an engine and aren’t flapping your arms, you are not really flying”.
I personally use the autopilot less, I prefer hand-flying in turbulence and below 10,000ft, but that is a personal choice to remain in practice
