Kelly (20 page)

One day we were visited at Lockheed by Assistant Air Force Secretary James Douglas and Gen. Clarence Irvine of the Air Force. Their question was, “How much stretch have you got in this thing, Kelly?”

“Let’s take a look at it,” I said. “Here’s the inboard view. You can see it’s totally liquid hydrogen from one end to the other except for a small cockpit up front.”

You do not put liquid hydrogen in nooks and crannies and odd-shaped tanks. The container has to be cylindrical—and very well insulated. With this airplane, we didn’t have the condition we’ve always had with other aircraft, both piston-powered and jets, where extra fuel could be added for a little more power and range. The Constellation gross weight, for example, doubled in its lifetime. We were able to do the same with most of the fighters, too. But with the liquid hydrogen airplane, once you set down the tank volume, that’s it. You could carry external tanks, but it would be difficult, and the airplane would carry added drag because of air resistance.

So the Secretary and the General turned to Perry Pratt, head of Pratt & Whitney engine design. “Maybe there’s something in the engine. Perry, how much stretch do you have in this engine?”

“Perhaps three to four percent in five years,” was the answer.

Overall, it wasn’t a very good forecast. We all agreed to cancel that effort without any more expenditure of funds. And had we proceeded, we would have run right into the energy crunch of the 1970s.

Coal slurries—finely-ground-up coal mixed with a light oil base and water—were a possible power source, injected into the engines as fuel. But the tiny coal cinders tended to ruin the turbine blades.

Boron compounds—in slurries—were tried, too. But these were difficult to use and plugged up the injector nozzles, not only in the engine but afterburners as well.

We decided to stick with liquid petroleum as fuel. It would have to be a very special fuel, though, for those operating
altitudes and temperatures—from – 90°F for midair refueling at high altitude to + 650°F in supersonic flight.

We put the problem to our old friend, Jimmy Doolittle, now a top executive with Shell Oil. That company had come up with what we called LF-1A, “Lockheed Lighter Fluid 1 for the U-2; and it was a good fuel for that airplane. Shell came through again, with cooperation from the Ashland and Monsanto companies and Pratt & Whitney, with a new chemical lubricant and fuel for the Mach 3-plus aircraft. This we call LF-2A. The Air Force has its own designations, of course.

Development of that fuel took a lot of doing. It was very expensive, but it’s an excellent fuel.

The fuel in this design also acts as an insulating element. The fuel tanks not only contain the fuel but are constructed to protect the landing gear. The gear retract up into the middle of the tanks. With radiant cooling to the fuel, the rubber tires are insulated against the very high temperatures the plane encounters during hour after hour of flight. And the plane doesn’t land with the tires ready to blow out.

In selection of structural materials for a Mach 3-plus airplane, aluminum automatically was ruled out. It would not withstand the ram-air temperatures of 800°F over the body of the plane. High-strength stainless steel alloys or titanium would be required for the basic structure. And high-temperature plastics would have to be developed for radomes, cockpits, and certain other areas.

Stainless steel actually is a better high-temperature material than titanium. But visiting the Lockheed-Georgia plant, which was building parts for the supersonic B-70 bomber, I saw what it took to make basic honeycomb panels for the fuselage: a “clean-room” environment—what was essentially a big pressurized airbag—with pressure locks for entrance and exit, everyone in white clean-room suits, and all the controls necessary to observe sterile conditions.

I reverted to my old Skunk Works axiom, “Keep it simple, stupid.” The more complexity, the more potential for problems. This is too sophisticated for the Skunk Works, I decided.
We’ll use the material we’ve worked on experimentally for ten years, the new advanced titanium in conventional structure.

The shape of the airplane itself was determined after a great many wind-tunnel and other tests. The result, head on, looks like a snake swallowing three mice. And for good reason. We added chines on the fuselage to get aerodynamic lift, among other things. We had some without them, but a terrific amount with them.

Before we got into high gear on production, we thought it advisable to build several test samples of the most complex sections: the nose and the basic wing structure.

The first wing section was a catastrophe. When we put it in a “hot box” to simulate high in-flight temperatures, it wrinkled up like an old dishrag. The solution was to divorce the skin panels from the wing spars in each direction and put corrugations and dimples in the skin—the wing surface. When the titanium got really hot, the corrugations merely deepened. I was accused of making a Mach 3 Ford Trimotor—that was made all of corrugated aluminum. But it was a very effective solution to a really difficult problem.

The nose section of the airplane presented other problems. We put it in the hot box to study cooling requirements for the pilot and the gear. We produced 6,000 parts; and of them fewer than ten percent were any good. The material was so brittle that if you dropped a piece on the floor it would shatter.

Obviously, we were doing something wrong. We queried Titanium Metals Corporation on why we had hydrogen embrittlement from our processes. They didn’t know. So we threw out our entire titanium processing system and replaced it with the same methods TMC used in making the original sheets and forgings at their factory.

After the initial shambles on the nose segment heat treat tests, we put into effect a quality-control program that I believe was and is unequaled anywhere. For every ten parts manufactured, we made three sample parts. These would be heat-treated and otherwise tested before any of the others of the batch would be put in storage for future use. One sample went
into a tensile strength test machine to find out how strong the material itself was. In the second, we made a short cut—about one-quarter inch long—and bent the sample at that cut around a form with a very small radius—as small as 32 times the thickness of the sheet—to see if it would crack. The third sample was used in case a re-heat-treat test was necessary. We didn’t want to throw away the whole batch needlessly; it was too darned expensive.

We could trace back to the mill and know the direction of the sheet rolling, and whether the part was cut with or against the grain. Before we would do all the expensive machining to cut landing gears from the huge heavy extrusions we would cut twelve samples, and unless everyone met the test we devised for them, we would not use that extrusion to make a landing gear. We’ve had no landing gear failures on the birds despite the hard landings that go with in-service flying.

There were times when I thought we were doing nothing but making test samples. But the test effort was worth it. By the early 1980s, we had made more than 13,000,000 titanium production parts for all of our Skunk Works airplanes and also for the Lockheed L-1011 commercial transport and the company’s big military cargo aircraft.

Titanium is such a rigid material that it cannot be shoved into place—as can some other metals—and therefore cut to less-exact tolerances. It must be tooled to fit. While this exact tooling is very expensive, it saves in the long run on scrap parts—of which there were almost none in production.

We had to invent a very large press that would shape titanium under very high temperatures—up to 1500°F and very high pressures.

The tough titanium actually is a very sensitive material to handle. Everything wants to poison it. We learned at an early date that we had to take cadmium-plated tools out of the mechanics’ tool boxes if they were going to work around the engine, because the cadmium would flake off enough to poison the bolts. After only one or two runs where they attained temperatures above 600°F, the bolt heads would just fall off. We
had to keep cadmium away from the titanium.

We found that the spotwelds on the wing panels failed very early in their test life when we built the panels in the summer, but if they were built in the winter they would last indefinitely. Analyzing all the processes, we discovered that in summer the water supply system for the city of Burbank was loaded with chlorine to reduce algae. When we washed the welds with pure water, there was no problem.

Special tools were required. When we first tried to drill the heat-treated B-120 titanium, a drill would be totally destroyed after about 17 rivet holes. Finally we found suitable drills developed in West Germany. Today we can drill more than 150 holes with a drill; and resharpened, it will drill another 150 holes.

We had to train thousands of people, not only our own, but Air Force mechanics and employees of our subcontractors and vendors—more than 300—in how to handle the machined parts. It’s difficult to get an old-time machinist to change his ways. He wants to discover on his own how to do something. So in the Skunk Works we put them in the experimental shop under the engineers’ direction and made them a party to developing the data. That always is a good tactic: involve the employee in the whole program as much as possible to arouse his interest and inspire his best performance.

One thing we learned in manufacturing the first Oxcart airplane was not to trust color coding. I had insisted on color codes for all wires and tubes and other connections, so that plumbing and other systems could not be installed incorrectly. Working with that many people, we discovered that ten percent were color blind. We’ve found a part bent over four inches to be connected incorrectly. We still color code, but we also use odd-shaped terminals that will fit only one way for those who can’t distinguish colors.

Materials and manufacturing were only part of the problem. There were the systems—hydraulic, electrical, and others.

Redundancy of systems has been a design requirement of mine ever since we added the auxiliary fuel pump to the F-80 after we lost Milo Burcham. The Constellation had it for the first
completely power-boosted controls. The Lockheed L-1011 airliner has it in triple and quadruple systems. The Blackbirds have triple redundancy. It is a safety factor that pilots especially appreciate. The cost, if you start with it in original design, is only a few tenths of a percent of the total. When you consider the value of the human lives and the vehicle this redundancy is protecting, the cost seems even smaller.

Hydraulic fluid was another special problem. First, of course, we surveyed all the suppliers to see if any of them had a high-temperature fluid, able to operate at above 600°F. One responded with literature on a fluid that worked at 960°F. I requested a sample immediately. It came in a canvas bag! That’s a funny way to ship hydraulic fluid, I thought. When I opened the package I found a white crystal.

Yes, it would operate at 960°F, but it was a solid at ordinary temperatures. You’d have to thaw out the hydraulic system with a blowtorch—not too useful for an airplane. So, the hydraulic fluid became a development project, too. The final product was a basic fluid developed by Pennsylvania State University, but with seven ingredients that we added so that it would withstand the temperatures and still function as a lubricant for the pumps and other hydraulic gear.

There were a few other little items—all important. Leather washers or rubber O-rings could not be used at those temperatures. Steel was the answer, giving no trouble at high or low temperatures. Fuel tank sealants to contain the fuel were another necessary development. While the airplane is not totally tight and will leak some on the hangar floor, the fuel has such a high ignition point that it is much safer than ordinary fuel.

Electrical problems alone threatened success of the project. We were not able to complete a flight on the Oxcart without some kind of failure due to the electrical system—which controlled the autopilot, flight control system, navigation system, and with electrical transducers even the hydraulic system.

At one time, 17 percent of our flights had to be cut short because we couldn’t measure oil pressure. We could not take a chance on burning out those very expensive engines. We had
to institute a cooling system for the oil temperature gauge.

Especially critical was the electrical transducer measuring air displacement related to proper positioning of the inlet spike. It was our toughest electrical problem.

We simply were not able to get the electrical system to work reliably under conditions of very high altitude, very high temperature, and very substantial vibration.

I personally spent six weeks at the test base working on this problem. The entire project was at stake.

Finally, we had to invent our own wire for the electrical system. We used high-temperature Kevlar wiring and just wrapped asbestos around it in the critically hot sections.

Special plastics were designed not only to withstand the high temperatures but to provide low radar return.

The Blackbirds take their name from the dark blue-black paint. The color was determined after tests for emissivity—heat emission from the hot airplane in flight. Emissivity can make a difference of 50 to 80 degrees in temperature on the aircraft, so it is a critically important item. Actually, the color of the Blackbirds becomes blue as temperatures increase at high speed and altitude.

We had to invent a special paint for the Air Force insignia. After just one hot flight, the red would turn brown and white become mottled. Getting the paint to adhere to the plane at all was another problem. We were getting little pockmarks on the painted high-temperature plastic that makes up 20 percent of the aircraft surface. We found that when fuel had been spilled, and temperatures of 550 to 600 degrees reached, little miniature explosions were occurring on the plastic skin. The paint had to be made fuel proof as well as rain proof.

The various payloads—cameras, very sophisticated electronic gear, navigation systems, inertial systems—all were the result of great effort in development.