Pilot-induced oscillations: are you a sinner or a victim?

You have probably seen this before: a GoPro video showing a pilot struggling with large inputs on the yoke, giving the throttle a hard time with either high thrust or idle power, and after a fair amount of time focused on that demanding approach, a smooth touchdown followed by a reassuring smile. On the title of the video, something mentioning a high crosswind component, and below, the comments saying that the pilot nailed it like a boss. Did he or she?

It is no secret that pilots like to hand fly—after all, that is what make them aviators. No one watched Top Gun during their childhood and started dreaming about monitoring an autopilot-flight director system for seven hours straight. So, although the industry has slowly but surely turned airline pilots into managers, to the extent that it is itself now concerned about their hand flying capabilities, we, on the other hand, kind of pardon the boring hours of paperwork and flight path monitoring for the sake of those couple minutes taming the beast.

On final
Wrestling the airplane to the runway—why?

Yet apparently some colleagues are taking it too seriously, and their thrilling performances are a great reminder to ourselves. After all, it requires a high level of self consciousness for one to know exactly how he or she is dealing with the flight controls. Recording yourself flying in video, instead, is a very useful tool for this matter, and although understandable, it is sad that most airlines do not allow their pilots to do so. Since you are very focused on flying, looking out the window and to the instruments, and your hands are responding almost automatically, to notice what movements you are actually doing is not that simple. Nevertheless, some people have recorded it, and after posting themselves on the internet, is time for us to look at them and try to figure out if we are doing the same unintentionally.

Imagine yourself flying, and while you are preparing for an approach, you encounter a challenging, windy condition ahead. The good side of it is that you are going to be more focused on the safe outcome of that approach, so the chances of you ruining your touchdown, for example, are greatly diminished. It is not uncommon to see basic mistakes being made in good weather with calm winds, since complacency tends to kick in and get us relaxed. But how do high winds affect our aircraft?

First of all, there are limitations we have to observe. Some are imposed by your airline policy, others by the insurance company policy, and besides those, for those pilots who are fortunate enough to fly their own aircraft, it is always advisable for them to have their own personal limits. When it comes to the manufacturers, usually the word used is “demonstrated” crosswind component. That is not exactly a limitation: anyone who has flown a Cessna 152 long enough knows it handles crosswinds perfectly well above twelve knots, if properly dealt with.

But having hard limits on your operations, either from your experience or from the operator that pays your bills, is healthy. In the airlines we are talking about crosswinds up to forty knots, and although such conditions would keep most light GA airplanes on the ramp, in major airports around the world it is just another day at the office—probably followed by a beer and some goods stories shared far from home.

Having said that, pilots from all types of aviation and all levels of experience fall into the pilot-induced oscillation (PIO) trap. From an F-22 Raptor test pilot to a Boeing 737 captain, not excluding several single engine piston students and instructors, the over control is just like the controls themselves: all over the place. But why does that happen, especially in high wind conditions?

Let’s start from the basics.

Every action has a corresponding reaction of same intensity but opposite direction: this is one of Newton’s laws. So this is pretty much the way an airplane, an inherently stable flying machine, goes up, down, and to whatever side it has to: by changing the control surface positions, leading to a reaction of the whole airframe in the air.

Now during a final approach, the objective is exactly the opposite: to remain on a constant flight path that will eventually take the airplane to a touchdown on a predetermined portion of the pavement ahead, no matter how dynamic is the atmosphere around it. But then there’s the wind. It can come from either direction, so a tailwind tends to make you float, to approach too fast, and pilots usually have fairly low limitations in order to keep us in the space available for us to stop upon landing.

Headwinds, on the other hand, are usually welcome, and since it is unlikely you’ll find any headwind capable of making your aircraft to fly backwards, there are no real limitations regarding them (other than for autoland systems). But, since high winds, even headwinds, normally change in intensity the closer you get to the ground and are often associated with turbulence, some kind of work on the power is going to be required for sure. As usual, the gold rule, “pitch for speed, power for altitude,” goes without saying.

Airline takeoff
Even airlines pilots can occasionally be accused of “bull riding.”

But it is with crosswinds that our bull riders are more prone to shine. And I say that from experience: I was one of them for a fair amount of flight hours, and it took an instructor’s advice to make me realize what I was doing wrong. Just like pilots in movies who do sharp hand movements and make confident statements using sunglasses to look cool, the same coming from a real pilot on a real flight had no use other than for dramatization. And we will get to that in a minute. First let’s review the three main crosswind landing techniques. Crab, decrab, and sideslip.

Landing in a crab means you are not correcting for the wind with the wings. Instead, you are flying wings level all the way to the ground, while your nose is pointing into the wind, relying on the natural reaction the airplane has when affected by any sort of crosswind. Although it might be easier, since all you need to do is keep flying in the direction of the runway and its aiming point without messing with the rudder while flying, this technique is not practical for all types of airplane. Narrow body aircraft can usually do it on wet runways with ease, and even on dry ones with very high crosswinds (if coupled with a sideslip). But the widebody jetliners are the ones that take more advantage of it, because its robust bogie-mounted main gear can withstand some serious crab upon touchdown, tilting the whole machine back to the runway orientation almost by magic, especially with the modern, fly-by-wire systems behind the scenes.

The decrab method consists of turning the crab approach into a sideslip during the flare, which is so quick and precise that it leaves no room for funny inputs by the pilot and works well with pretty much any airplane. This is true especially when ground clearance by the engines or wings might be a factor, even at small bank angles.

But then we have the pure sideslip—the method of excellence for airplanes like narrow body jetliners (which would end up in the grass if crabbing into a dry runway) or single engine pistons (which, without this technique, would ruin their tires painfully as they touched down in a decent crosswind).

The sideslip method has a beauty all its own. It might not be as elegant as the crab when seen from outside, but it requires a decent amount of coordination on the flight controls by both human and machines—yes, automated landing systems use the sideslip, and this is the main reason why they have relatively low crosswind limits. The Boeing 787, for example, can handle up to 25 knot crosswinds, down to the roll-out, even on one engine. How amazing is that?

But humans can land in much more severe crosswinds, simply because we can add some crab into it and, of course, we have the capacity to react in a satisfactory way to quick changes that the computer would probably find too puzzling. So this is the perfect—although not only—scenario. You are approaching in a crosswind and choose to do it with a sideslip. Since, as mentioned before, turbulence is often associated with high winds, now you find yourself having to adjust to the changing movements of the airplane every tenth of a second, and that requires from you all your hard-earned skills. Suddenly, you are fighting the universe like there was no tomorrow, and either you are doing it thinking it is quite the right way to do it, or you are not even noticing the hard time you are giving to the airplane. Where does that come from?

Crosswind landing
To crab or to slip, that is the question.

In Portuguese, we have a very ironic expression to define it: “vento de cabine,” or in a direct translation, “cockpit wind.” Yes, you got it right: most of your corrections on the control were not even necessary in the first place, and could even be only a response to some overcorrection you did just before. That is why it is technically classified as pilot-induced oscillation. Most of the movements the airplane is making are a direct result of the pilot’s inputs. And if those inputs are not needed for any practical reason, then they are nothing but the pilot fighting himself, using the airplane as the battlefield.

How silly can it get? Well, a bit more. You see, even highly experienced, fly-by-wire aircraft pilots do it here and there, and many of these airplane types are able to distinguish the pilot’s inputs from the wind effect, thus correcting the latter to make our life easier. So, if you start correcting something that three flight computers already did, well, then it gets really embarrassing.

Some people got the bad habit during basic and even advanced training, from instructors who used to do it and made it look right. Others had this tendency naturally, increased by self confidence, and were never properly addressed by anyone they’ve flown with. The fact is, that is an obviously wrong thing to do. As you make large and quick inputs in a small airplane, you are exposing the airframe to loads it might not have been designed for. And if you are flying a big jet, then the whole inertia involved in airplanes that can have the area of an Olympic pool or even a city block, makes these quick opposite direction control inputs ineffective altogether—not to mention what your passengers are going to experience in the back.

So, if you have the chance to record yourself flying during a gusty approach, do it. Then analyze honestly your performance and come up with ways of improving it, if needed. If you can’t place a GoPro on your operator’s $250 million airplane, that’s also fine: the next time you fly into a windy destination, pay attention to the way you are acting on the controls. Do it gently, with small amplitude inputs, and wait for the airplane to react before you make your next move. Most of the adverse airplane displacement imposed by the wind is momentary, and it is not uncommon to end up where you were half a second before without doing anything—especially if you have a fly-by-wire system in normal mode assisting you.

One thing is for sure: there is absolutely no need to deal with the stick and rudder as if you were playing the drums.

24 Comments

  • Totally agree. I see so many pilots “fight” the airplane during approach and landing with most the control inputs unnecessary. I just chuckle to myself when I see this. When I was a sim instructor, I would sometimes set the winds to calm, but tell the crew the winds were strong and gusty with a good crosswind. Many of them would struggle to get the airplane to the ground, creating their own “crosswinds” and control problems. Some pilots just make it a lot harder than it really is.

    • That is a good catch, Mark! PIOs are like having an extra dose of medicine. If it is too much, will harm you instead of healing. Thank for your impressions!

  • I think there must be some middle ground to this. My flight instructor taught me this way, smooth gentle inputs, and I was all over the runway and it wasn’t until Rod Machado taught me to “buck up for the rodeo” that I was able to make a more precise landing in gusty crosswinds.

  • I don’t consider my approach stabilized until my control movements are at a minimum. And if it’s not stabilized; it’s power forward, positive rate, reset the trim and clean her up.
    Only times I’ve used large control movements was from wake turbulence early on. I no longer except “squeeze ‘em in approaches”.

  • This is a common problem with new student pilots. It usually takes a few hours to convince them that the stability and inertia of the airplane work better than their jiggling of the controls.

    • With me happened something funny. I was well taught from the start. When I first flew the jets, though, I had an instructor who would do just those exaggerated movements, and was picky on me for not doing them. So, after a while I got the bad habit I didn’t have before. It took another captain, during a windy approach, months later, to notice what I was doing. I got the message: from there on, I never needed large/quick inputs again.

  • I must say “Been there done that” On my initial IOE on the B727 after flying C-46’s and DC-7’s. On my second approach, going after it, I was working those flight controls, all of a sudden the Capt grabbed the yoke and shouted ” stop it, you are creating moderate turbulence with all this input”. Of course the 727 with their double acting spoilers really did roll us around if you used more that 10 degree input which isn’t hard.
    I NEVER forgot that approach. From then on realizing planes didn’t need all that input. I will say however you do need more input below 100 ft then at 500ft with a strong crosswind.

    • Well noticed, Bob! I never flew the beautiful 727, but even the 737 has clear trouble with the flight spoilers if you bank too much. The non-fly by wire airplanes are more demanding, of course, but even them have their own time on reacting to the environment, and our input must meet this time.

  • Read the March/April 2014 FAA Flight Safety Bulletin page 9 for explanation of the physical inputs to control by humans.
    Just because its published in that publication does not mean the same organization reads it or uses it!!!

  • While flying F-4s at Eglin, I bought an Ercoupe for personal fun. Those “castered” landing gear made most crosswind landings a piece of cake. I also didn’t have a lot of rudder action, since there weren’t any!

  • What about a good ol’ full workout in one of the squirrelier taildraggers that seem to be getting harder to find? I know of one operator who wouldn’t hire a pilot to fly cargo in twin Cessnas without a tailwheel endorsement.

    • Hey, Dan! I am a believer that no pilot is truly better than other: he/she is just specialized in something different. So, if I was to hire someone to fly a twin, a multi-engine certificate would make the big difference. Taildraggers make you good at… taildraggers. F-18s make you good at F-18s… and so on. Although every experience counts and is relevant, not every experience is essential. Thank you for the comment.

  • Rarely is a crosswind smooth and steady. They are usually gusty and generally decreasing in strength as you get closer to the ground. On a bad day they can be just as strong and gusty low down.They are generally a crosswind because the runways have been aligned to suit the most constant direction. I have found that , after being totally terrified as a student 40 years ago , an attitude of “ I can control this “ with constant small control inputs generally worked. You can always go round . ( and I have )

    • Precisely that, Graham. There’s a video on Facebook these days of a 747 approach… each input is like moving a house rooftop up and down in half a second. The inertia of the whole thing don’t even notice what you did.

  • The good thing of control inputs on final (100 ft or less), is when your rudder and ailerons are at full deflection, you know early enough to go around. You have exceeded your aircraft’s safe structural design limits- unless the tires and landing gear are 4-wheel drive rated and the wings are good for cartwheels.

Leave a Reply

Your email address will not be published. Required fields are marked *