Safety Pilot: Deja vu: 50 years of mishaps

March 1, 2008

Bruce Landsberg was named executive director of ASF in 1992.

In looking back at nearly 50 years of general aviation accident reporting, what is surprising is how few surprises there are. In GA, the reason we’re crashing has not changed since when The AOPA Pilot first rolled off the presses in 1958. Human nature doesn’t change much and accidents in most areas of life follow certain patterns. In the home, stairways, carpets, and ladders should cause some trepidation, as they always have. In cars, failure to yield right of way, speeding and following too closely remain popular accident triggers—not much change here despite technical improvements in the vehicles. The nature of mishaps doesn’t change unless the underlying technology causes a fundamental shift. And there are only so many ways to crash airplanes; most involve ground contact in other than landing configuration. There are some nuances but, we, as a group, aren’t too original. There are also well known solutions to all these problems.

Data for this comparison came from the FAA General Aviation Accidents database—1958 and the AOPA Air Safety Foundation’s 2007 Joseph T. Nall Report. It’s difficult to precisely equate the FAA’s 1958 accident categories to today’s NTSB description, so consider this commentary as a general comparison. Starting with total accidents, we’re doing much better in 2006, the latest year available, with 1,319 accidents compared to 4,135 in 1958. That’s a drop of more than two thirds. There’s no attempt to identify the accident rate by estimating flight hours in the early years, but it’s a reasonable guess that we flew much less than today since there were fewer GA aircraft. Fatal accidents numbered 273 in 2006 compared to 328 in 1958, roughly a 17-percent decline. (Previous comment regarding rate applies but the fatality numbers have fallen much more slowly.)

Despite significant technical improvements in aircraft, the nature of dangerous flight hasn’t changed if one is prone to risk taking. “Pilot struck or flew into an obstacle” (128 in 1958) is essentially the same problem in 2006 with 94 maneuvering mishaps. “Failure to maintain airspeed” still has the same consequences as it did back then. Some improvements in aerodynamics have made stalls more docile, but in the hands of an inattentive, ignorant, or overaggressive pilot the outcome is as it’s always been. Impromptu aerobatics in non-aerobatic aircraft has always been a bad idea.

One area where there has been marked improvement is in landings. Ground loops and noseovers have largely been eliminated in nosewheel aircraft. Tailwheel airplanes predominated 50 years ago and they regularly cracked up on landing—1,001 times to be exact. In 2006 the total number of landing mishaps, not just groundloops and noseovers, was 392. Moving the third wheel forward had a positive effect on the geometry of the gear and made the aircraft much less prone to groundlooping. The level of skill required dropped significantly and safety advanced. That said, we still have way too many off-runway excursions as some pilots still have difficulty with directional control or are unable to judge the flare correctly.

Aircraft that require a high level of skill and considerable training experience have more mishaps, thus the genius of moving the tailwheel to the front. The FAA formally changed pilot requirements for tailwheel aircraft in a FAR 61.31 in 1991 when it required those who had not yet obtained tailwheel experience to receive special training. It added an endorsement and specified currency in tailwheel aircraft to carry passengers, a nod to the more difficult nature of tailwheel ground operations.

While we’re on the topic of landings, the FAA noted that in 1958 there were 868 cases where the landing gear collapsed or was inadvertently retracted. Out of that group there were 30 fatal accidents. That’s just more than 3.5 percent. In 2006, there were 392 landing accidents from “all causes” but only eight involved fatalities or, put another way, about a 60-percent drop in fatal landing accidents compared to the early years. That could be because of better design, better emergency and rescue services, or something else.

Accidents attributed to powerplant failures in GA have dropped significantly (410 in 1958) even though the basic design of our engines has changed little. With a fundamental shift in engine technology, the current number of 99 accidents because of engine and propeller failure could drop lower. Airframe and components were identified as causal in 294 accidents in 1958 compared to 124 in 2006.

There are a couple of subtle messages in all these statistics. First, it is most likely that we as pilots will be the source of a mishap. Secondly, the hardware is about four times as reliable as humans and, although the machines do a very good job, they still have to be maintained. As the fleet ages this becomes more critical. By the way, humans always fare worse than hardware as the source of failure because you can re-engineer the failing part but you can’t re-engineer the human—in most cases. Education, while highly beneficial in certain circumstances, will always take second place to technology.

Weather claimed a lot of pilots in 1958. There were 109 fatal accidents, or better than two per week. The FAA’s document does not specify exactly what happened, but I’ll speculate that VFR into instrument meteorological conditions (IMC) was the leading problem area, just as it was in 2006. There were 32 fatal accidents with nearly 75 percent in single-engine aircraft because of a bad decision. In twin-engine aircraft, interestingly enough, a completely different picture emerged in 2006. There was one VFR into IMC, one loss of control by an IFR pilot, three icing accidents, and four entanglements with thunderstorms. When you consider the relatively small number of multiengine aircraft compared to singles in 1958, that leads to two hypotheses. These aircraft are likely being flown in worse weather, and twin pilots are pushing into to some very nasty conditions beyond what either they or their aircraft can handle. It’s all about limitations, people.

One area that bears watching in the future is the number of fatal descent and approach accidents, which was up significantly in 2006 to 41. My estimation is that many were under IFR. This was a category that barely existed in 1958 in GA since there were few IFR-equipped GA aircraft or pilots in the early years. The jump in numbers from 25 in 2005 may reflect all the high performance technologically advanced aircraft—or TAA—that are being sold.

These airplanes are used for transportation, which means they are being exposed to more approaches, which also explains at least some of the increase. It also argues strongly for pilots who are flying instrument approaches to shape up lest they and their passengers ship out early.

So we have a mixed message. Aircraft are better today but not yet benign and the vagaries of weather and the result of poor decisions continue to cause accidents. As in the rest of society, safety statistics have improved and generally continue to move in the right direction, albeit slowly. Flying is as safe as we choose to make it, and if that responsibility is taken casually, bad things can happen—just as they did 50 years ago. As for our part, the AOPA Air Safety Foundation will continue to offer first-class training materials and seminars to more pilots than ever before, just as we have for the past 58 years.

See this year’s Joseph T. Nall Report and the Online Safety Center on the Web site.