Not what you might think

This happens pretty frequently, at least to those who spend a lot of time following accident reports. It’s not just because certain types of causes are hard to distinguish before the factual investigation’s complete. It may, for example, seem clear that a loss of thrust precipitated the sequence, but unless an aircraft with five hours of endurance was aloft for five and a half, or the fuselage was found covered in oil, the reason for that loss may take some time to determine. It could be a physical failure of the engine, the fuel system, or ignition; fuel exhaustion, starvation, or contamination; carburetor icing; or foreign material blocking the air intake. It’s possible that no specific cause will ever be found. This happens dozens of times each year.

Want some examples? When an amateur-built Sonex crashed into a California vineyard in March 2009, initial speculation centered on the 75-year-old pilot’s age, health, and limited experience—he’d received his sport pilot certificate just three months before. The probable cause turned out to be that he failed to tighten the locknuts after adjusting the engine’s valve clearances, which allowed an exhaust valve to slip out of contact with its connecting rod.

The following month, a Piper Arrow made a forced landing in a Florida back yard after bypassing a planned fuel stop. However, the NTSB found “sufficient fuel in both tanks” at the accident site—and a broken cylinder stud with its nut still attached on the runway at their point of departure. Insufficient torque on the nut during an engine overhaul 131 hours earlier led to the fracture of the stud and subsequent separation of the number three cylinder from the crankcase.

Early reporting on the in-flight break-up of a Pilatus PC-12 in Florida focused on the presence of thunderstorms across its planned route of flight. In fact, the airplane never got close to convective activity. While being vectored around the precipitation, the airplane’s autopilot abruptly disconnected. Rather than flying by hand, the pilot became so distracted with troubleshooting the autopilot that he failed to notice the turboprop entering a nose-down unusual attitude. It reached an airspeed 175 knots above its maximum maneuvering speed before he pulled it apart trying to recover.

Speculation about the attitudes or actions of pilots who are no longer able to speak for themselves is especially problematic. Yes, the NTSB may eventually draw some inferences, but they’ll base them on more information than you have. When a Cirrus SR22 entering a Maryland traffic pattern collided with a Robinson R44 departing on a rental check-out, the airport community was strongly inclined to believe that the airplane pilot must have descended early. In fact, both aircraft were equally far off their altitudes, with the airplane 200 feet low and the helicopter 200 feet high. Did a lively conversation with another commercial pilot detract from the helicopter instructor’s customary watchfulness and adherence to procedure, or did the renter just blow through his altitude? There’s no way to know. Likewise, the initial report that the pilot of a Beech B36TC that crashed attempting an ILS approach into Amarillo, Texas at 1 a.m. wasn’t instrument-rated immediately suggested spatial disorientation. It turned out, however, that the pilot had been signed off for his instrument checkride (which had been postponed after he’d failed the written test) and had flown three more IFR flights earlier that day—perhaps with the assistance of the Bonanza’s very capable three-axis autopilot.

Enough of aviation is counterintuitive that it shouldn’t come as a surprise that first impressions often prove wrong. Until the facts are in it’s better, as Benjamin Franklin noted, “to keep your mouth closed and be thought a fool than open it and remove all doubt.”