May 2007 Volume 50 / Number 5
Take it off
AOPA Air Safety Foundation Executive Director Bruce Landsberg is always pleasantly surprised when the aircraft flies.
It's not often that pilots reject a takeoff. In multiengine aircraft we obsess over rejected takeoffs or engine failures, but in single-engine aircraft a common attitude is that there's not much the pilot can do, so why worry? During my early flights I marveled that the airplane flew at all, and really didn't expect it to get airborne. That's not a bad way to think — expect the worst and be pleasantly surprised when flight actually occurs.
The AOPA Air Safety Foundation accident database shows that the approximate number of fatal takeoff accidents in 2001 was 49 and by 2005 it had risen to 62, although the FAA estimated that flight hours decreased about 9 percent. That's going in the wrong direction. Over that five-year period 232 fatal accidents occurred in this phase of flight. I say "approximate" because there is some arbitrariness about where takeoff ends and climb begins. Likewise, the NTSB might categorize something as a takeoff accident when ASF might list it as a fuel-mismanagement accident, so the numbers are not exact; but it's a reasonable estimate and includes enough crashes to warrant some attention.
When the engine quits
In single-engine aircraft the powerplant is a traditional concern. Clyde Cessna once said, "If the engine stops for any reason, you are due to tumble, and that's all there is to it!" In Kansas, where Clyde and company built thousands of aircraft, there was often a big wheat field right off the end of the runway to cradle aircraft when engines quit. In more populous areas, land demand puts buildings, highways, and power lines on potential off-airport landing sites. In many cases the planning, or lack thereof, by the zoning authorities is truly astounding; they apparently have great confidence in our ability to consistently make a successful takeoff.
From what ASF could verify, there were 18 mechanical engine failures with an additional 43 in which the engine stopped for "undetermined" reasons. That works out to either 7 percent or 26 percent, respectively, of the total number of takeoff mishaps. As is typical of the pilot-aircraft relationship, the remaining majority of accidents were something pilots inflicted upon themselves. Although the investigators could find no reason in 43 cases for the engine to quit, given the millions of takeoffs successfully performed each year in single-engine piston aircraft, it's not something we should obsess about.
We won't spend much time discussing the impossible turn after the engine stops. The general advice is to avoid a tight turn back to the airport until reaching maneuvering altitude. The lack of altitude and airspeed usually leads to stalls and spins with bad outcomes. It's much better to hit something soft and cheap that's out ahead of you and to maintain a flyable angle of attack all the way to the ground (see "Push: Your First Move When the Engine Fails," November 2006 Pilot). A tight turn vastly complicates that. Once established on the crosswind leg, there's a decent shot at getting back to the airport, if not the departure runway.
How long is my runway?
Adequate runway for takeoff is essential. Profound thought! In a number of accidents, the pilot didn't do the math and wound up going off the end or stalling by attempting the impossible climb. In a recent accident, a homebuilt Rutan Defiant attempted takeoff on a short runway. The NTSB noted, "According to the owner's manual, the airplane was a heavy wing-loading aircraft that required longer runways than typical general aviation twin-engine airplanes." The estimated ground roll for takeoff was 1,920 feet and the available runway was 2,150 feet with a 485-foot overrun. No problem, right? The wreckage was found 2,867 feet from the start of the runway. With a warning like that and not much experience in the aircraft, one has to wonder why this pilot stacked the cards so badly against himself.
This is a good place to discuss accelerate-stop distance for single-engine aircraft. That's the distance it takes to get up to flying speed, decide that you've had a change of heart, and abort the takeoff with maximum braking. That's a great option to have if the takeoff isn't working. It's standard procedure for multiengine equipment, but you don't hear much about it with singles. It's not a number that you'll see in the single-engine aircraft's operating manual or pilot's operating handbook (POH), but we'd sure like to see the manufacturers start publishing it.
Here's how to make an approximate calculation: Take the takeoff ground roll under ambient conditions and add it to the ground roll for landing. It's a good idea to add in some recognition and reaction time as well, unless you're really good at split-second decisions. Consider three seconds as a minimum. That's not very much and at typical rotation speeds for a light fixed-gear single, that adds another 300 feet. Here are some hypothetical ambient conditions: sea level, dry paved runway, no wind, and a temperature of 20 degrees Celsius. For an older-model Cessna 172, this works out to 835 feet for runup to rotation speed, 300 feet for decision time to reject, and another 530 feet to smoke the brakes and bring it to a stop. That's 1,665 feet and with an additional 10-percent margin allowing for older engines and brakes, not to mention pilot technique, it works out to about 1,800 feet.
Play with the numbers for your aircraft a few times and by rounding up to even hundreds for the various components, the math is simpler and you'll quickly estimate if the runway distance gives you a fighting chance at not bending the aircraft if the engine isn't up to snuff before liftoff. If you're a little slower in reaction, as many of us are, allow five seconds and 500 feet, which then yields a 2,000-foot accelerate stop distance in the above example. For additional consideration — grass or any runway contamination, such as water or ice, will greatly extend the stopping segment. Finally, how good are your brakes? Rest assured that the flight-test aircraft had perfect brakes but ours are not likely to perform at that level.
Obstacles in your way
Speaking of engine malfunctions, frequently the engine gives ample warning that it isn't having a good day. There were more than a few instances where the magnetos were not up to spec or the witnesses heard the engine backfire. Takeoff roll was sluggish and yet the pilot persisted. This is not exactly the time to be persistent. If the engine doesn't give you 100 percent, taxi back in and get it checked. Wishful thinking has yet to get an aircraft airborne.
Sometimes the famous 50-foot obstacle upon which all takeoff calculations are based isn't 50 feet. Early in my career at Cessna Aircraft Company in Wichita I observed the takeoff performance testing on the then-new Cessna 172RG. In Kansas, as previously mentioned, trees are in short supply so the engineer used an optical measuring device that showed where the aircraft was exactly 50 feet above the ground. I came away with a newfound respect for calculating takeoff distances because the margin to clear the obstacle was minimal — way closer than I would ever want to be to a real obstacle, and to top that, the obstacle was imaginary. Would you care to wager how much more distance the test pilots would factor in with their families on board if the obstacle were real?
You've heard all the warnings on how the tests are conducted on a perfect aircraft, with a new engine developing full power, on a perfectly flat runway with a perfectly competent test pilot. Here's an easy way to add some margin for clearing the not-so-imaginary obstacle with the less-than-perfect aircraft and pilot. Use the POH distance under ambient conditions to clear the obstacle and add 50 percent. Under the conditions stated earlier, the Cessna 172 needs 1,490 feet to just fly over the obstacle with everything going your way. Add a margin of 750 feet, or make the math easy and use 1,500 feet plus 800 feet. With 2,300 feet as the minimum runway distance and using our calculated accelerate stop distance of 1,800 or 2,000 feet, now we're thinking like the professionals and building in some margins for this most critical phase of flight.
There are many more aspects to takeoff accidents to discuss another time but let me focus on a big one. An engine stoppage because of fuel starvation or exhaustion during the takeoff phase of flight is a real "aw-shucks!" — often with fatal consequences. Sometimes fuel starvation sneaks up on the pilot who's been working in the pattern or hopping rides. Everything is normal and perhaps the pilot is running an abbreviated checklist. However, "fuel on proper tank" (with fuel in it) is not something to be overlooked, because there is very little time to sort through options and get a restart.
Takeoff expectancy is every bit as dangerous as landing expectancy. The fact that it has always worked before does not serve pilots well for a critically quick response mindset should there be a self-inflicted or mechanically inflicted glitch.
If you'd like to see the takeoff accident history for your aircraft make and model, visit the AOPA Air Safety Foundation Web site and click on Accident Analysis, select ASF Accident Database.