Aftermath: For Want of a Nail

Pilot training often underestimates the element of surprise.

Aftermath Artwork
Accident analysis that goes behind and beyond the NTSB report.Flying

The pilot’s loss of airplane control,” said the ­National Transportation Safety Board, “was due to his diverted ­attention to the canopy opening in flight.”

At what point the canopy opened ­completely is not clear, but several witnesses described the airplane rocking sharply from side to side and oscillating sharply in pitch during the takeoff roll, movements that suggest the pilot was trying to close the canopy while continuing his takeoff. The airplane became airborne, made a tight 180 to the right, and flew parallel to the runway at a low altitude. “I could ­clearly see his head,” a witness recalled, “and there was no glass cover above him anymore.” The airplane then turned tightly back toward the ­runway, banking close to 90 degrees, and began to ­descend. A wingtip caught the ground; the pilot was killed on impact.

The runway was 10,000 feet long. He began his roll at an intersection but still had 7,500 feet ahead of him — enough runway to be able to stop, secure the canopy and resume the takeoff under the same clearance. Why did he not do that? Evidently, he thought it would be easy to close and lock the canopy with his left hand — the throttle hand — while his right remained on the sidestick.

“I could ­clearly see his head, and there was no glass cover above him anymore.”

The airplane was a homebuilt Rutan VariEze originally licensed in 1992. The pilot, a minister who used his airplane to visit different congregations, had purchased it eight months before the accident. He was at least its fifth owner.

The VariEze, like other Rutan canard ­designs, is naturally incapable of stalling. When the stick is held full back, the nose rises and falls gently as flow on the canard surface separates and reattaches. The main wing remains unstalled, with full roll control. This is an important point. In a conventionally configured airplane, dropping out of a tight turn to final would point to a stall-spin; in a VariEze, which cannot stall, it only implies that the pilot failed to control the flight path.

The VariEze’s canopy is hinged along the right upper edge of the fuselage tub. It ­rotates upward to open; a cable behind the pilot’s headrest keeps it from touching the right wing. Nearly all VariEzes had one-piece bubble ­canopies; the accident airplane, however, had a two-piece variant, with separate sections covering the front and rear seats. In both versions, the windshield is an integral part of the canopy.

The latching system consists of three linked hooks that pull the canopy snugly downward. The foremost hook, which the pilot operates, engages a latch that keeps it from being accidentally released.

Around 1979, a fatal accident similar to this one occurred. The pilot forgot to latch the ­canopy, it flew open, and he lost control of the airplane. The problem was not aerodynamic; the VariEze is quite controllable with the canopy open. In fact, one early owner who used his airplane for air-to-air photography would deliberately open his canopy in flight and considered flying that way “a no-brainer.” But he probably used goggles, since, without a windshield, the pilot has no protection against a very strong wind.

Even if he wears glasses, his eyes will fill with tears, blurring his vision. This combined with startle and disorientation is a major reason for losing control. Startle and disorientation, even without the blast of wind, are sometimes enough. Certified production airplanes have crashed after a door popped open, even though the change in their behavior due to the open door is barely detectable.

Almost nothing that can happen in flight requires such urgent, immediate and complete attention from the pilot that he cannot continue to fly.

The first concern of a pilot confronting any unexpected inflight event should be to fly the airplane. Almost nothing that can happen in flight requires such urgent, immediate and complete attention from the pilot that he cannot continue to fly.

After the 1979 accident, RAF published in its Canard Pusher newsletter a mandatory modification to the ­canopy design. It required builders to install a backup latch, similar in principle to the hood latch on a car. If the canopy were unsecured, it would rise just an inch or two before the backup latch caught it. The backup latch came to be called the “Rice catch” after the unfortunate pilot who lost his life.

It appears, from photographs taken minutes before the minister’s crash, that his VariEze did not have the Rice catch.

It appears, from photographs taken minutes before the minister’s crash, that his VariEze did not have the Rice catch.

If the VariEze had been a certified airplane, the FAA would have published an airworthiness directive, and the modification would likely have been made ­mandatory before further flight. Any A&P inspecting the airplane would have detected a failure to comply. But no such mechanism exists in the world of homebuilt airplanes, where each owner becomes an official inspector for his airplane — a one-trick airframe mechanic.

This particular VariEze had changed hands ­several times. If it was originally accompanied by a pile of ancient Canard Pushers, they had probably gone astray at some point. Or perhaps the original builder had bought the plans or a partially completed airplane from someone else, long after the trail of the Rice catch had grown cold.

The accident pilot, far down the chain of ownership, had probably never heard of the backup catch. If he even asked himself why there was not some system to keep an unlatched canopy from flying open, he may have inferred that the canopy had no tendency to open in flight or, if it did, that it would be a simple matter to close it again. As one RAF publication wrote: “Builders of ­Experimental aircraft are, in general, careful to do inspections and check all newsletter ­notices. Those who operate but did not build their aircraft are ­generally at a much greater risk, since they ­often ignore notices or are not experienced at inspection and repair.”

The steps leading to this fatal ­accident were four. First, the canopy lacked a backup catch. Second, the pilot failed to lock the canopy before taking the runway. Third, he did not abort his takeoff but tried to correct the problem on the fly. Finally, once airborne, he did not focus on keeping the airplane under control and climbing to a safe altitude.

Four steps — four opportunities — and none of them taken.