Skunk Works

The biggest delay was in my bailiwick. We had contracted the inlet control mechanism that would move and position the cones to Hamilton Standard, which was also doing the fuel control system for Pratt & Whitney. The trouble was that the pneumatic inlet controls they devised were not responding quickly enough. We had spent $18 million to develop this system, but after more than a year the problem was unsolved. Finally, I took the matter to Kelly. I said, “Kelly, I think we’ve got to cut our losses and find someone else to get the job done.”

He cussed and agreed. We went to a company called Air Research and they developed an electronic control that saved our bacon. And they did it in less than a year. Meanwhile Pratt & Whitney was struggling with a slew of problems that were putting them further and further behind schedule. Dick Bissell and his assistant, John Parangosky, watched in anguish as our delays and costs mounted. In pique, Parangosky had begun referring to the P & W engine as the “Macy’s engine,” and complained to Bill Brown, their program manager, rather unfairly, “If we gave as much money to R. H. Macy’s, they could build that engine in time for Christmas.” By mid-1960, the agency decided to crack a mean whip. Kelly was called on the carpet and told he had to accept a CIA engineer bird-dogging in his shop and looking over all shoulders. Kelly blew up. He knew the agency was only trying to cover their own asses. But Parangosky, whom Kelly liked and trusted, flew in from Washington and warned him that if he refused the request there was a grave likelihood that the CIA would cancel the contract entirely.

Kelly fumed. “No, John! I’m not gonna have one of your spies poking into my business. Bissell promised me you guys would keep hands off and let me do this thing my way just like the U-2.”

“Kelly, be reasonable. We won’t get in your way. We just want someone here you can trust and we can too.”

He suggested a very bright and able engineer on the CIA’s payroll named Norm Nelson, whom Kelly had known from World War II days, and both liked and respected. Kelly sulked but ultimately surrendered. “Well,” he said, “I’ll let in Norm Nelson, but not another goddam person. You got that? Besides, you tell Nelson he can have a desk, a phone, but no chair. I expect him in the shop, not sitting on his fat duff.”

Nelson, who arrived in the spring of 1960, became the first outsider ever allowed a place inside Kelly’s realm. He gave Norm a free hand and actually took suggestions because he respected Norm’s judgment. We all knew that he reported directly to Bissell, and he knew that we kept him out of certain meetings and didn’t let him in on everything. But Norm was sharp. I recall one meeting in 1961, when Kelly told Norm that the agency had given us an additional $20 million to develop wing tanks on the Blackbird to extend its range. Norm did some quick calculations and figured out we would extend the airplane’s range only eighty miles. “Fooling around with wing tanks at this point will be more trouble than it’s worth,” Norm insisted. Kelly said, “You’re probably right.” Kelly Johnson sent back the $20 million that afternoon.

But our biggest problem was about as easy to conceal from Norm as a pregnant pachyderm on top of a flagpole. Norm Nelson came aboard just as we were starting to build a mock-up of the fuselage-cockpit section, which would contain more than six thousand parts, for heat testing inside an oven. To our horror, we discovered that the titanium we were trying to use was as brittle as glass. When one of the workers dropped a piece off his bench, it shattered in a dozen pieces. The trouble was diagnosed as poor quality-control procedures in the manufacturer’s heat-treatment process—a problem that caused endless delays, forcing us to reject 95 percent of the titanium delivered and set up a rigorous quality-control procedure.

For an outfit that detested red tape, we now found ourselves wallowing in bureaucratic procedures. We sample-tested for brittleness three out of every ten batches of titanium received and kept detailed records of millions of individual titanium parts. We could trace each part back to the original mill pour, so if a part went bad later on, we could immediately replace other parts from that same batch before trouble developed.