Free Kelly by Clarence L. Johnson Page B

wondered if we weren’t overlooking something. I had seen the tab control, the ball bearing that is supposed to keep the tab in proper position as the rudder turns left or right. The bearing had been broken and there were no balls in it. The whole center race—the track the ball bearings ride in—was gone. And so was the tab—the movable trailing edge of the tail rudder.
    Back at the plant, I convinced Hibbard and others that we should run a tunnel test on a full-scale vertical tail and find out what conditions would cause flutter. We built our own wind tunnel in 1939, the first sophisticated one in private industry. But at that time we were able to use the Guggenheim wind tunnel at Cal Tech under the distinguished Dr. Theodore von Kármán and Dr. Clark B. Millikan. The test section was a cylinder ten feet in diameter, obviously a very difficult space in which to work when changing models, but the tunnel had enough capability to exceed by far the speed at which the Northwest aircraft was cruising when it was lost.
    That tail wouldn’t flutter in the wind tunnel no matter what we did. But then we disconnected the tab, simulating thebroken bearing, and immediately the rudder blew off the tail assembly. We went ahead with the lead balances decreed by the CAA; but the real cause of the accident in my opinion was that someone, a mechanic on the production line or in the airline’s overhaul shops had not held the bearing properly when he adjusted the tab setting and had cracked the race but not broken it completely. The break occurred later in rough air and caused immediate violent flutter. I had never seen such violent flutter as with the simulated condition in our wind tunnel.
    We still had to conduct 50 hours of flying in icing and rough weather to satisfy CAA conditions. And it had to be in the same air corridor where the plane had been lost. So Headle, Holoubek, and I headed for Minnesota. We stayed on the ground when the weather was nice and everyone else was flying, and took off in the worst of it.
    We would fly into the roughest air we could find to prove stability of the aircraft, and in the worst icing conditions to prove the new non-icing carburetor. We’d have two to four inches of ice on other parts of the airplane, but the engines kept running. On one of our most exciting flights, we collected so much ice in just four minutes that with both engines on full power our indicated airspeed was just 90 miles an hour. And we landed with full power on.
    I was so impressed with the rapidity with which ice can build up and its severe effect on aerodynamics and control that I was prompted to write one of my first technical papers on the subject for the benefit of others. (“Wing Loading, Icing and Associated Aspects of Modern Transport Design,” Journal of the Aeronautical Sciences, December 1940). To this day, the only thing I fear more than ice is hail. A separate icing wind tunnel later was built as another research facility at Lockheed. Pilots who haven’t experienced it do not realize that it takes damned little ice to cause a horrible crash.
    Another little detail we had to prove was that the control cables would not go so slack because of the low temperatures that they would allow flutter under certain conditions. To do this we had to measure elevator cable tension, and the onlyplace to reach the cable was by removing the toilet and reaching down through that space to attach a tensiometer. This was Holoubek’s job as inspector. One day while he was checking this instrumentation, we hit a particularly severe bump. I still can see his feet sticking up through the opening as he yelled for help to get out of there.
    The Model 14 had so much power with those new Wright-Cyclone engines that this actually produced a problem. With so much of the inboard wing in the propeller slip stream, it was almost impossible with power on in flight to stall the middle of the wing. So the wing, if it stalled, would stall—that is,