Thanks for the list! I would guess that 5 of the 6 would have been less likely to happen if they had the Chelton.
I know what all of you think about the training issues and the “antilock brake syndrome” where accident rates actually went up. I agree with this analysis and worry very much about it, but I am not sure if there is anything we can do about this short of changing the nature of flight training to it’s core.
Maybe it’s just wishful thinking, but I am hoping that the Chelton will change the dynamic of flight in VMC. If I’m wrong and it only makes things worse then I don’t think there is much of a hope of solving this problem.
The answer is an emphatic YES. You must never hand fly to help the autopilot. Let’s say you’re in a right turn being done by the autopilot. You think the turn is too steep and you manually apply left aileron to correct. The autopilot senses a load on the aircraft that is countering its instruction to turn and reacts by increasing the force for a right turn. Same with pitch control. If you are in altitude hold and pull back on the stick the nose will want to pitch up. The autopilot will react by trying to lower the nose to maintain level.
This is a function of all autopilots and can get you in real trouble. On occasion if you try to overpower the autopilot you will cause the autopilot to attempt to apply more trim than it can and it will disconnect in a very out of trim condition. That can lead to immediate loss of control.
Some autopilots have a function called Control Wheel Steering (CWS) that allows you to momentarily disconnect the autopilot, change the plane’s attitude and the immediately reengage. The Cirrus autopilots do NOT have this function.
Either fly manually or use the autopilot. Don’t try to do both simultaneously.
I’m not Neil, but I’ll share my one inadvertent entry into IMC. Was cruising along under a low deck, with 3-4 mile visibility. Had it on autopilot and buried my head in the cockpit studying either the map or GPS, looked up and I was in the clouds. I had started my IFR training, but only had about 12-13 hrs. hood time. That was sufficient to execute a 180 to exit, but it wouldn’t have helped me much beyond that.
Of course the entry into IMC was inadvertant, on one occasion between converging cloud layers, and on the second in a rain shower, with declining visibility, but in both cases I executed a 180. I also know one should never try to overide the autopilot.
My point was, perhaps this could be a factor in some of the incidents we are discussing here. If a pilot is scud running with the autopilot engaged, inadvertantly enters IMC, and attempts a turn to exit the IMC, is there a chance the aircraft responds in an unexpected manner.
In spite of “sign offs” and recurrency training, how much don’t we know about each of our own abilities/judgment?
You make an excellent point. In life in general, I think that one indicator of “maturity” is a more realistic understanding of our own abilities/judgment/vulnerability etc – hence the common perception of the recklessness of youth, etc.
Perhaps the other extreme is represented by the very elderly, who understand their weaknesses very well, and frustrate those of us who can’t understand why anyone would want to drive at 20 mph on a major highway!
Back to your point - it’s certainly valid for me. Looking back at my early flying, I count myself lucky to have survived it; although at the time, I felt that I had ample tools in my judgment arsenal to keep myself safe. Now I know a great many things I did not know then, and if anything, I wonder what major lessons I have yet to learn.
If a pilot is scud running with the autopilot engaged, inadvertantly enters IMC, and attempts a turn to exit the IMC, is there a chance the aircraft responds in an unexpected manner.
Neil,
It depends on what you mean by “unexpected manner”. As Jerry pointed out, it is TO BE EXPECTED for the autopilot to “fight” you if you try to override it by pushing against it. Fly the plane, or let the autopilot fly the plane, but don’t allow both to fly the plane.
My point is… control may become difficult if a pilot tries to fly in parallel with the autopilot. Is this unexpected? Well, maybe, to a pilot who doesn’t know this. Autopilot or no, any VFR pilot in IMC should “expect the unexpected”.
Neil, I don’t want to sound harsh but you said that 1) you inadvertently entered IMC when you shouldn’t have and 2) you know you shouldn’t try to override the autopilot. Yet you ask if the autopilot’s behavior when the pilot attempts to override it can be a factor in some of the accidents involving VMC into IMC.
I would argue, strongly, that the reason for VFR into IMC accidents is the pilot’s failure to take action so that IMC is not “inadvertently entered” and that the pilot’s misuse of the autopilot might be an aggravating factor.
As Mike said, the autopilot reacted exactly as it was supposed to. The problem is that the pilots can and do abdicate PIC responsibility to the autopilot and that can only eventually lead to disaster, especially when and if the pilot doesn’t understand how the autopilot is supposed to work.
My question would be, “Why would you want to fight the autopilot?” The best thing to do would be to set the heading bug to the current heading, change the autopilot to heading mode, and turn the heading bug to the desired new heading (e.g., 180 degrees away from the current heading) Even for an instrument-rated pilot, this is safer and easier than transitioning to instruments and hand-flying, with the autopilot on or off.
We all make mistakes. The trick to surviving them is to not make additional mistakes in trying to compensate for the first one.
All this assumes that your airplane has a heading bug, of course.
My point is… control may become difficult if a pilot tries to fly in parallel with the autopilot. Is this unexpected? Well, maybe, to a pilot who doesn’t know this. Autopilot or no, any VFR pilot in IMC should “expect the unexpected”.
I’ve done this a few times (flew in “parallel” with the autopilot, not flown VFR into IMC).
For some reason, sometimes my brain registers turning off altitude hold as turning off the A/P. For example, if ATC gives me a fairly long descent (i.e. descend and maintain 3000 when I’m at 8000), I’ll disengage altitude hold but leave the A/P engaged.
Problem is, by the time I reach 3000, for some reason I think I’ve already disengaged the A/P, and try to fly by hand for a while until I realize it feels way out of trim, and then realize my mistake!
I must have done this at least 2-3 times since owning the plane! Oops!
I’m not sure, but probably to do with the fact that the '22 has an additional pitch servo (i.e more than just the “trim pitch” that we have on the '20). In the '20 arrangement, the autopilot manages pitch with trim. In the '22, it has a more direct way of controling pitch, but it can also manage the trim as needed. This means that as the pilot, you have more to overcome if you’re fighting it. It also accounts for the fact that '22s are more often out of trim in the pitch axis when you disconnect them, vs. a '20.
As I said, I’m not sure – that’s my understanding of the arrangement. Perhaps a 22 owner can confirm or correct.
Mike: That’s the arrangement all right, in fact, the autopilot pitch servo failed in my SR22 in mid-training. It is mounted under the baggage compartment floor and is made by STec. I can’t say I have done much fighting with either autopilot.
The most I do is when I expect slight turbulence when IMC is to “follow” the autopilot with a sort of light grip just to make sure it is doing the right thing. However, anytime I don’t think it is doing the right thing or turbulence gets more than slight, I am ready with the disconnect button.