On January 18, 1911, Eugene Ely landed his Curtiss pusher airplane aboard the USS Pennsylvania anchored in San Francisco Bay -- the first successful shipboard landing of an aircraft, and the first ever using a tailhook system designed and built by circus performer and aviator Hugh Robinson.
After the first trap Ely told a reporter: "It was easy enough. I think the trick could be successfully turned nine times out of ten."
In the case of the Joint Strike Fighter's initial aircraft carrier developmental testing, Ely was off a bit -- by six, to be exact. The first 10 times the F-35 tried to perform an arrested landing -- with experienced test pilots at the controls -- the airplane only caught a wire three times.
You don't need to be a tailhooker to figure out that that percentage won't work out in the fleet. Jets come back from missions usually with a handful of looks at the deck at most, and if a pilot puts his craft in the wires, he should have confidence he's going to stop.
So what was going on?
The main contributing factor to this high incidence of in-the-wires bolters was that the JSF only has 11 feet between the main mounts and the hook point. That's more than 10 feet less than any other tailhook airplane the Navy has ever flown from carriers. Follow-on video analysis showed that the relatively short wheel-to-hook distance didn't give the arresting wire time to bounce away from the flight deck after the wheels hit it at touchdown, and that caused the tailhook -- with a standard-shaped hook point -- to drag across the wires instead of catching them.
Arresting wires don't lay directly on the flight deck; they're elevated by curved pieces of metal known as "shives." So the engineers' first thought was to raise the shives so that the hook might have a better chance of catching. But the Navy wasn't keen on tackling a ship modification when the system worked fine for every other airplane, so the engineers looked at changing the JSF hook point instead.
The result is a tailhook with a sharper point that sources tell us appears to have solved the problem.
And so we have another data point around why we do flight testing and why it takes so long for airplanes to reach the fleet . . . besides the convoluted DoD procurement process.
And for all of the computer modelling wizardry and 500-pound brains that populate the military's systems commands, you never know how something is going to work until you actually try it.