June 2014- We all learned to do stalls (however reluctantly) in our primary flight training, and demonstrated them to the satisfaction of an examiner in order to get our licenses, and we’re all aware of the consequences of an inadvertent stall near the ground. What else is there to say?
Just one thing: when did you last do a full stall?
I’m having a hard time remembering, and that worries me because I just recently learned that my stall training—and most probably yours—didn’t include one very important real-world item: the first thing to do in stall recovery is to turn off the autopilot!
Let me go a little further: the first thing to do if the airplane is not responding properly to control inputs is to turn off the autopilot!
Why do I bring this up? Over the past few years, there has been a rash of stories about experienced airline pilots making major (in some cases, fatal) mistakes when dealing with stalls. The most famous is probably Colgan Air flight 3407, in which the flight crew of an iced-up commuter airliner responded to a stall by pulling—rather than pushing—on the yoke.
Another famous accident was Asiana Airlines Flight 214, where it turned out that none of the four pilots (including three captains) on the flight deck had their hands on the throttles as the airplane slowed—and eventually stalled, landing short of the runway.
“So what,” you may say. “Those were airline pilots flying complicated hardware. It’s not relevant to us piston drivers. We do a lot more hand-flying.”
That’s what I thought at first, too, but an article in the March 24, 2014 issue of Aviation Week & Space Technology made me reconsider. In response to a series of stall-related crashes, the FAA has decided to mandate stall training for air transport pilots (ATPs).
This creates a problem. Full stalls aren’t practical (and in most cases, aren’t allowed) in transport-category airplanes. Stall awareness training is done using full-motion simulators, which weren’t built with enough range of motion to accurately simulate a full stall.
However, there are a variety of things the simulators could do to get the idea across. To judge whether any of those things might be adequate, the FAA got no less than 45 experienced ATP volunteers to try out different ways of simulating something less than a full stall in a modified simulator.
Now here’s the interesting bit. What the FAA did was to have each volunteer go through two hand-flown stall demonstrations; basically the same thing you and I did in flight training (and maybe have been asked to do during a BFR), but they added something else: a surprise stall.
The surprise happened during a simulated ILS approach. The simulator was programmed with a whopping tailwind that built up so quickly that the airplane stalled (a fair representation of what can happen in severe wind shear as the result of a thunderstorm gust front).
Just 10 of the 45 volunteers reacted correctly—by disconnecting the autopilot, lowering the nose and adding power to recover.
Note that sequence: an airliner flying an ILS does so with the autopilot in approach mode, automatically following the localizer and glideslope. In a severe wind shear situation, airspeed drops (in this case so much so that the airplane stalls); but left to itself the autopilot will try to keep on the glideslope, which just makes matters worse. The pilot needs to hit the disconnect button—it’s usually on the control yoke—and then recover the aircraft in the way we were all taught.
Why does this matter to you?
Does your airplane have an autopilot? I’m betting that if you use your airplane for more than just the odd hop around the patch, it probably does. I’ve owned (or been a partner in) four airplanes so far, and all but the first had an autopilot.
Simple wing-levelers aren’t as much of a worry in this situation, but if you’ve got an airplane with altitude hold, or an autopilot with approach mode, you can have the same situation hit you as the FAA tossed those volunteers—and it will be just as important that you disconnect.
Stalls on approach aren’t the only thing to be concerned about. Another story in the very same issue of Aviation Week discussed the loss of a KC-135 Stratotanker when the pilot tried to compensate for a malfunctioning yaw damper using the rudder pedals. The resulting oscillations completely tore the tail off the airplane.
Again, you may ask, “Why does this matter to us prop pilots?” I did a quick search of the NTSB accident database and turned up a case where the pilot of a PA-46 experienced a sudden unexpected yaw which was only overcome when he turned his yaw damper off.
He was lucky, and got down in one piece—but the aerodynamic forces were strong enough to create severe airframe damage. And the PA-46 isn’t the only single with a yaw damper.
Typing “autopilot malfunction” into the NTSB database returns a scary list of cases where pilots have lost control of the airplane either because the autopilot disconnected suddenly, or because the pilot got distracted trying to deal with an autopilot issue.
Another point to think about concerning the issue of a surprise stall: as piston airplane pilots, none of us should be flying close enough to a thunderstorm for the FAA gust front scenario to be a real risk.
Then again, have you ever tried to land just ahead of storm? That’s a setup for getting into a gust front.
After reading the Aviation Week articles, I decided to try an experiment. Getting into N4696K and flying around looking for a gust front wasn’t practical, and I don’t have access to a full-motion simulator, but I do have Microsoft Flight Simulator on my desktop PC.
I started it up, selected User-Defined Weather with a visibility of one mile, and set up my virtual piston single for the ILS approach at my home airport. Just outside the marker, I paused the sim, then went into Custom Weather Options and set three wind layers: 20 knots on the nose from 30,000 feet, with a mild wind gradient; 30 knots from the side at 1,000 feet with a moderate wind gradient; and 40 knots on the tail at 500 feet with a strong wind gradient.
I hit pause again, got the virtual airplane on the glideslope at the marker, set approach mode on the virtual autopilot, and waited to see what would happen.
I’m very happy I wasn’t trying this in a real airplane—by the time I got the autopilot turned off and got control, the attitude indicator was showing nearly all blue and then nearly all brown. I did get it under control and made an impressively fast low pass on the virtual airport before I quit—with some decidedly non-virtual sweat on my brow!
Yes, that was a lot of wind—but so is a thunderstorm gust front. And that’s not the only way to have a surprise stall. Thinking it over, while I’m not sure of the last time I deliberately stalled an airplane, I have vivid memories of three nasty surprises—twice in the vicinity of mountains (probably as a result of mountain wave action) and one on my way back into my home airport at night after a very long day of cross-country flying.
Fortunately, I was hand-flying all three times. If I’d been flying an airplane with a more advanced cockpit, things might have been different.
Next time I go out for hood work, before my usual three approaches I’m going to head for the practice area and make sure I can at least handle power-off and -on stalls—that I’m expecting—from under the hood. – JDR
John D. Ruley is an instrument-rated pilot and freelance writer. He holds a master’s degree from the University of North Dakota Space Studies program (space.edu). Until recently he was a volunteer pilot with ligainternational.org and angelflight.org, two charities which operate medical missions in northwest Mexico and provide medical patient transport, respectively.Send questions or comments to editor@www.piperflyer.com.
RESOURCES
Further reading
“The Anatomy of an Accident: Colgan Air Flight 3407,” by Robert L. Sumwalt. Published by the National Transportation Safety Board, Feb. 12, 2010.
https://www.ntsb.gov/doclib/speeches/sumwalt/SCAA-100212.pdf
“Stall Tales,” by John Croft. Aviation Week & Space Technology, March 24, 2014.
“Hard Lessons,” by Tony Osborne. Aviation Week & Space Technology, March 24, 2014.
NTSB Aviation Accident query page
ntsb.gov/aviationquery
