Reading the (excellent) new book Into the Black, about the flight test history of the Space Shuttle, I was intrigued by this line:
“Engle was a low-gain pilot. Like Charles Lindbergh or Chuck Yeager, he barely moved the stick, anticipating the need to do so and making small, necessary corrections in plenty of time. His inputs were smooth and progressive, never snatching at the controls.”
Rowland White, Into the Black.
The author is talking about Joe Engle, aeronautical engineer, test pilot, and NASA astronaut. He was scheduled to land on the Moon, a dream dashed by budget cuts, but ended up going back into space by flight testing the Space Shuttle. He’s a no-kidding ‘pilot’s pilot’ who loves flying anything with a lifting surface. And apparently someone who made small smooth movements of the control stick: A low-gain pilot.
It’s a concept that really does connect a pilot skill in all kinds of flying, all kinds of planes. From the simple Piper Cub stick, to the Airbus video game stick, to the complex F-16 and the Space Shuttle. I’ve seen it first-hand in master pilots several times over the years. They are ahead of the plane. They have a light touch. They make gentle early corrections.
I’ve seen it flying sailplanes with Jason Stephens, multiple-time glider aerobatic champion and owner of the famous Arizona Soaring flight school. Talking about high-performance soaring, he told me every little un-needed control input causes a little drag, and then needs a correcting control input that causes a little more drag. One of his ‘secrets’ of maximum performance is to be a low-gain pilot.
I saw it in the simulator with a USAF fighter and display pilot who flew combat in Vietnam and instructed in the T-37, F-4 and F-15. He later became an airline check airman in the B-727, F-100 and B-777. He repeated “be smooth” and “be small.” The light touch on the stick, the smooth movements, were such a part of his flying that his military call-sign was Stroker.
Legendary British test pilot John Farley, in his 2008 book A View from the Hover: My Life in Aviation, writes,
“Over-control is a common problem with learning to fly, almost regardless of the task but with experience we get better at relaxing, better at trimming, better at letting it fly itself for a bit and then coaxing it back to the desired state. In fact better at becoming a low gain (relaxed) pilot rather than being a high gain (overactive) one. Airplanes take time to respond and it is a waste of time to oscillate controls.”
Slow down and listen again to the words he uses: Relaxing. Fly itself. Coaxing.
It’s not just planes that respond well to low-gain inputs. It’s common to racing cars as well. You’ll see it watching Jackie Stewart talk about going fast in F1: “Monte Carlo should be driven smoothly and quietly”. And the one time I got to fly a helicopter, I spent every second I was holding the stick saying “small pressures, small pressures”.
The opposite to low-gain piloting has a phrase I’ve had thrown at me in tailwheel training and Airbus simulators: ‘stirring the pot’. The hapless pilot in the flare, not quite sure of what’s happening or how to fix it, ends up making loads of fast, seemingly random, inputs. The stick goes into continuous motion. Faster and bigger. Stirring the pot. Sometimes it works. Sometimes it doesn’t.
I used to tell primary students that the yoke came from Cessna completely smooth — it was nervous students gripping too tight that wore the finger grips on the back side. I thought it was funny. But there was a serious point to be made. You can’t make small smooth control inputs if you are gripping the yoke too hard. Relax. Slow down. Breathe.
Think ahead of the airplane, so you see early, when a small correction is all that is needed. It’s part of being one with the wing. If you are reacting late it’s hard to be low-gain. And if you are getting desperate in the flare, it’s easy to start stirring the pot.
After 18,000 hours I’m still trying to loosen my grip. Still trying to make early small smooth inputs. Still telling myself ‘don’t stir the pot’. Still trying to fly like Lindbergh, Yeager, Engle, Farley, Stroker, and Jason.
Still trying to be a low-gain pilot.
A very beautiful article, which I’m sure everyone could take a little note of.
I am intrigued by this post too! I think this ought to be discussed a bit to clarify if today’s young pilot/engineer should indeed strive to be “low-gain”. I write this with utmost respect for Dave and the pilots mentioned in the post. My admiration for The Inner Art of Airmanship is what compels me to comment.
My perspective is that of an engineer and derives from academics and work in aircraft control systems. I am also a pilot and have my share of sloppy landings under my belt. This comment is a bit long as I am going into some pedagogical detail, with examples. How would a controls engineer (EE, CS, ME, Aero, etc.) interpret the terms used?
In control theory, a simple form of control is “Response = Gain x Error”. This is called proportional control. At the simplest level, ‘Gain’ could be a constant number. There are two other terms that come to mind:
– Dead-band: This is the small magnitude of error, over which no response is made.
– Control update frequency: This is how often the error is sensed and the response updated. Quadcopter and inverted pendulum controllers perform a lot better with 100 Hz updates versus 10 Hz updates.
In the extreme case of low gain, Gain equals zero. Hence, there is no response to error. The aircraft’s inherent stability (or lack thereof) becomes apparent as it flies with the stick fixed, regardless of error. This isn’t desired. Too low a gain can lead to a sloppy system. A very high value of Gain indeed causes over-corrections and could even lead to instability. Typically, Gain, as I have used the term, has an optimal value. Any higher or lower gain will lead to more energy usage.
I often end up not realizing that there’s an error. Then I correct vigorously because the situation is about to get out of hand. This sounds like having a dead-band, and then responding too late with a high gain multiplied to a large error. The airplane groans!
What might work better is to:
– Strive to eliminate your dead-band. A distracted, tired, or task saturated pilot finds it more difficult to notice the error as it germinates. An attentive pilot notices the slightest deviation. He/she pays attention to the far end of the runway or the horizon. Minor errors are perceived by their senses.
– Have the discipline to correct errors as soon as they are perceived. A focused pilot is able to make timely corrections when the error is small. Often these corrections are very small, but I wonder if that means the Gain is low. Probably not. Would a high gain with minuscule error and a lower gain with large errors work better?
– Pay attention and be one with the aircraft so the frequency of response updates keeps up with the frequency of error and disturbances. The pilot’s workload increases as the frequency increases, just like CPU usage!
– To fly gracefully. The relaxed grip and demeanor leads to smooth inputs even though the brain has updated commands at a fast rate. Some slew-rate and filtering is added to the response by the arm that moves the stick. Practice trains the mind and body.
There are caveats and there are extensions to the simplified control equation I have mentioned. My comment is to enable the readers to think along the lines of the engineers who worked on these legendary aircraft or, at least, help today’s student pilots relate the ideas to what they read in textbooks — even if they conclude that I am dead wrong! Indeed, none of this makes it to the tarmac as I go bouncing around stirring the mess!
So, are you a low-gain pilot?
Fantastic thoughts Chimney, many thanks for sharing.
The engineering/test pilot part of designing the stick and it’s control sensitivity is a huge subject. The many differences between, say, an A320 with its normal law compared to a powerful acro prop plane. It is a little odd how quickly we humans adapt to widely different required control pressures and movements switching between different aircraft, how quick the feedback loops self-correct.
I think you really are on to something with “Control update frequency” and being attentive enough to spot corrections needed when they are very small. So they are ‘low-gain’ because that’s all the correction needed when spotted early. Powerful insight.
Thanks again for your kind words, and I’ll see you in the sky . . .
Dave
So I’ll offer a specific case where this low gain theory doesn’t seem to work for me personally. When making an approach to landing on short final right up until completion of the landing flare and making ground contact…during gusty crosswind conditions my hands and feet are making continuous small and not necessarily smooth corrections. Relative to how I normally fly which could be considered “low gain” it’s pretty jerky and almost comical in comparison. So it’s still fairly small corrections but about double or triple the normal rate.
I also instruct flying and I encourage my students to “ do whatever is necessary” for landings, to not be afraid of using the controls. The idea is for the plane to fly the pilot and not the other way around. I can’t talk fast enough to tell the the control inputs required, I just reassure them if they are overdoing it or doing it incorrectly I’ll take over in time. Finally I was once caught flat footed so to speak encountering massive wake turbulence in a light twin following a relatively large passenger jet, again on short final, which required some very fancy foot (and hand) work. I guess the common theme is it was near the ground where the margin for error was rapidly decreasing. That’s it. 🙂