It had started as a run out to the Pilot’s Lounge at the virtual airport to check on plans for our annual end-of-the-year bacchanal and ended up with me wandering into a chance to learn from my betters. Sandy, our airline pilot, lost a friend of hers in a post-takeoff crash a few years back. On this day, she brought in the accident report because she wanted to show it to Old Hack, our resident curmudgeon and occasional purveyor of aeronautical common sense.According to the NTSB, Sandy’s friend had suffered an engine failure at what witnesses estimated to be about 300-400 feet over the departure end of the runway. Despite the relatively flat ground straight ahead, the pilot chose to make a turn of about 135 degrees toward another runway that had its approach end adjacent to the departure end of the takeoff runway. The cold truth of the matter was that Sandy’s friend, a pilot with about 3000 hours, stalled the airplane during the turn. The airplane impacted well nose-down. Sandy’s friend died in the accident. His passengers, though badly injured, survived.Sandy had brought the report, along with a couple of academic pieces on whether to turn back for the airport following an engine failure after takeoff, as she wanted to confront Old Hack on the issue. Hack had recently made the statement that a pilot who knew the airplane could successfully turn back after takeoff from an engine failure at 350 feet AGL and make a landing on the departure runway. Sandy thought Hack was all wet and she was ready for a fight on the subject. I stood well clear as Sandy opened up on Hack, figuring that staying out of the crossfire was the wise thing to do.
To Sandy’s astonishment, Old Hack would not engage. He said, “I think you’re right. If the engine craps out at 350 or 400 above the ground, almost every pilot would try to turn back to the runway. I’m convinced that nearly every one would die in the attempt. However, I think you could pull it off. But I don’t think you’d be stupid enough to try. You fly for a living. I’ve flown with you in that Citabria of yours and listened to you brief your takeoff every time. You push that throttle forward knowing in your gut that the engine could very well quit, so you won’t waste two or 12 seconds denying that a bad thing is happening … you’ll take action right away. You know where the wind is from, you know what’s off the end and to the sides of the runway, you never accept an intersection takeoff, so you have the maximum runway available, and you have decided what to do in the event of a problem before you started moving. You’ve planned things out and you’ll act. On top of that, you’ve maneuvered airplanes down low, something very, very few pilots have done. You are aware how the world looks different when turning at 200 feet than it does at 800; how the horizon seems higher and how you seem to be going much faster because the ground is closer and is whizzing by the windows.”Sandy stood staring at Hack. She’d been all geared up for a good donnybrook with her ancient friend and he wouldn’t give her the pleasure. I thought it was going to spoil her whole afternoon.Hack continued, “Sandy, in that Super Cruiser of mine, I know for a fact that I can easily make a 180 and then a little more to line up with the runway if I start from 500 feet AGL on a good day, with good visibility so I have a clear horizon. Without a good horizon, all bets are off. I’ve also spent a lot of time down low, scud running, and I’ve practiced that turn. I’m even confident I could do it from 400 feet AGL and make the runway if I’ve got about a 10 knot headwind. I probably could do it at 350 feet if I have been warned that the engine is going to quit and took action instantly. The problem is, I’ve read some of those papers on the ideal bank angle for a return to the runway and what I get from them is that the question of the altitude needed — as well as appropriate bank angle for the wind direction and velocity — is complex. Right then, you don’t have time for complex; what you need is a lot of simple. There just isn’t time to do equations and ponder variables when you have to act instantly. So, I’ve set myself a hard floor of 500 feet AGL for a turn-back after takeoff. The problem is, I’m not sure I’ve worked the problem through as well as I could and I prodded you on this the other day hoping you’d come in here and we could hash this out.”It was about then that the three or four other folks in the room relaxed. I figured we were going to have a chance to learn something, so we sat down and shut up. I took notes.Sandy had the results of some tests that had been done in simulators and in the real world with different types of airplanes. There were all sorts of turn-back techniques explored, from pitching up until out of speed and pivoting to holding a constant, very-steep bank angle toward the prevailing crosswind, to using some combination of flaps, ailerons and rudder to enter a steep turn, to even an incipient spin, rotate a half turn, shove the nose down and flare. The common conclusion was that the turn back to the runway required a change of direction more than 180 degrees because it offset you from the runway, even if turning into any crosswind; the ideal angle of bank was on the order of 45 to 55 degrees with the speed just barely above stall and then the nose had to be stuffed down hard to get enough speed to flare and land. The folks organizing the tests found that pilots almost never could make the turn successfully the first time, even if the victim knew the engine failure were coming. With training, in the simulator, things improved radically, so long as the pilots were ready for the engine failure. Once the element of chance again entered the equation, the rate of success dropped off badly.
Cosmic Pass/Fail Exam
What struck me as scary about the turn-back scenario was that the accident history and testing exercises demonstrated that there just wasn’t a happy medium or an “average” score; one in which the landing wasn’t necessarily beautiful, and the airplane got dinged up a bit, yet everyone walked away. When you look at something like crosswind-landing accident stats, you see that there are one heck of a lot of loss-of-control accidents in crosswind landings, but most result in a bent airplane and minor injuries at most. For the turn-back-after-takeoff-engine-failure exercise, the result is a cosmic pass/fail exam. If you do it right, you tend to land more or less on the runway, under control. However, if you fail, you jump right into the dead column. You stall the airplane and hit out of control, at a steep angle, which just doesn’t make for a very survivable impact sequence. The trouble is, according to the studies, pilots who try this exercise the first time almost always drop into the “fail” column.
It’s Different Down Low
I listened as Sandy and Hack talked and went through the materials Sandy had brought with her. Having spent a certain portion of my life flying very low and making turns near stall speed in crop dusters of marginal power and lots of drag, I listened attentively when Hack raised the point about pilots who practice this sequence up at altitude. He said that they pull the power to idle, count out loud to let three or four seconds go by — as they should to simulate the real-life fact that most pilots freeze up for about that long when the engine fails — and then roll into a steep turn and try to position for an imaginary runway. They discover that they can do it with an altitude loss of roughly 400 feet, but they also get amazed by the fact that the VSI absolutely bottoms during part of the turn. Wrapping around in a steep turn near stall speed and then pushing the nose down to get enough speed back to flare, the rate of descent rate will exceed 1,000 fpm for part of the exercise. Those performing the test always seem to conclude that they could pull it off if it happens for real.Those who practice at altitude don’t understand how the world looks different when trying to do all of the described aggressive maneuvering at 400 feet AGL. Until a pilot has spent time tossing an airplane around down low, it is hard to explain how the perception is different. The horizon seems higher. The ground is nearer — dramatically so. In a turn, the ground is a powerfully close blur of color that is a stunningly integral part of your peripheral vision: It’s right there, bigger than life and, by gawd, it’s going by fast. That’s something one doesn’t experience in normal flight, even when maneuvering steeply at altitude. Up there, the ground is a more remote, abstract concept and it seems to be moving slowly. Down low, with the ground moving fast, and with the groundspeed increasing while the airplane holds a constant speed near the stall and turns from upwind to downwind, even a pilot with low-altitude experience feels the very powerful sensation of the groundspeed increase and tends to unconsciously pull back on the yoke to keep the speed under control (and, sadly, despite all training, to try to keep from going down). What makes it even worse is that, when the nose must be pushed down hard to accelerate to get speed for the flare, all the pilot sees is a windshield full of ground. It takes a lot of training to accept that visual picture long enough to get enough speed to avoid a stall. When experienced for the first time under the massive stress of an engine failure, it is no surprise that the end is almost preordained even for high-time pilots: a stall, with the future existence of the pilot and passengers telescoped to mere seconds.
Get The Nose Down Now
Sandy raised another salient point: how very fast the airplane decelerates in climb attitude once the engine quits. I’ve watched it with students and pilots on flight reviews. When the engine stops, there is an initial sense of disbelief and reluctance to accept that the nose has to go down, to make a big pitch-change toward the earth to maintain flying speed, as well as stunned incredulity at just how fast the speed goes away. A friend of mine who took his initial dual in a Stearman told me that, during takeoff engine-failure practice, his instructor required his students to push the stick forward hard enough to generate negative Gs and shove the occupants hard against the seatbelt. It was necessary on an airplane that is drag incarnate such as the Stearman, but only marginally less so on a more streamlined machine. Remember that, on some airplanes Vx (best angle of climb speed) is set so as to clear an obstacle but is notfast enough to allow the pilot to lower the nose quickly enough to have enough speed to flare and land without collapsing the landing gear. Even climbing at Vy or Vy plus 20, you have to actively work to get the nose down to keep from stalling the airplane. And there is that huge psychological resistance to lowering the nose aggressively, or at all, when close to the ground.On top of everything else, Hack pointed out that to really hold the ideal speed in the turn back to the runway, one needs an angle-of-attack indicator. Few of our small airplanes have them, so we are forced to rely on the stall warner, which should be blaring throughout the turn, if we are to do it right. When was the last time you practiced holding speed by aural indications?
Have An Emergency? Why Not Create Another?
Sandy brought it home to me when she said that we all are taught to land straight ahead if the engine quits on takeoff and we’re below 500 feet AGL. (Sure, we can make a small turn to avoid hitting big stuff.) The good thing is that history and accident studies show that landing straight ahead is the way to most effectively reduce the risk of dying if the engine quits on takeoff. She also said that a steep turn, made near stall speed, near the ground, power off, is something she truly considers to be an emergency procedure. Therefore, what kind of foolish sort reacts to the emergency of an engine failure on takeoff by ignoring training and intentionally creating another emergency, one he or she has not practiced? Nature has little tolerance for outright stupidity. Maybe, just maybe, that’s why such a high percentage of those who try to turn back after a takeoff engine failure die in the attempt. The sad part from a Darwin-Award perspective is that many take innocent passengers to the grave with them.
Hack was very interested in Sandy’s technique of briefing all of her takeoffs. As he had never had crew-resource training, the briefing concept was new to him when he first flew with her a few years ago. He immediately came to like the idea. Sandy explained that she generally carried over the procedure from the airline, modified for a single-engine airplane. She assumed that something would go wrong on the takeoff and she wanted to be ready for it. She also wanted to monitor each step of the takeoff to assure that things were going right and so she might catch a problem before it became serious. That’s why she briefed the runway direction to match the compass, the obstructions around the runway and the wind direction, where the airplane should break ground if it were making power, and what she would do if the engine failed on the roll, in the air below 500 feet and in the air above 500 feet. She told me that if the engine failed below 500 feet, she had long ago accepted that the airplane would be a write off and her goal was to impact wings-level, slowly, without stalling, and to fly the airplane all the way through the crash so as to maximize the chance of occupant survival. She also said she had adequate insurance on the airplane and, with falling airplane values, she consoled herself with the idea she might be able to afford an even better airplane after the accident. On the takeoff roll, she called out engine rpm at full throttle to assure the engine was making power and also called out when the airspeed needle came off the peg. Otherwise, she kept her head outside the cockpit and watched to see that the airplane came off the ground about where it should.In the air, her engine-failure rule-of-thumb was to land on the remaining runway if possible, even if it meant rolling off the end slowly, because the airplane was pretty crashworthy and could accept a 30-knot impact with a fence better than hitting the ground in a steep turn, nose down, while stalled. Below 500 feet AGL, she figured on landing more or less straight ahead, depending on what obstructions were present, with a turn up to 90 degrees allowed once over 300 feet. At 500 feet, she normally started her turn to crosswind and, once in that turn, said she would most likely try to go back to the airport if the engine quit.Sandy’s approach made sense to me. I’ve had engine failures and I know for a fact that I spent a little time being surprised each time it happened. Having a simple procedure that has been thought out ahead of time means that the chance of surviving an engine failure down low is good. I’ll leave the academic exercises of steep turns down low to the ivory-tower types.See you next month.
Want to read more from Rick Durden? Check out the rest of his columns.