Skunk Works

I had a staff of three (by Skunk Works standards that was almost an empire). On the air-conditioning team, I had two engineers to help design the internal cooling system to safeguard the camera bays and the avionics and landing gear systems. The cockpit environment also presented a unique problem: without effective and fail-safe cooling the pilot could bake a cake in his lap. And as head thermodynamicist, that problem fell in my lap.

We designed the cockpit air-conditioning to bleed air off the engine compressor and dump it through a fuel air cooler, then through an expansion turbine, into the cabin at a frigid minus 40 degrees F, which lowered the ovenlike 200-degree cockpit to a balmy Southern California beach day. Developing these systems took us a year of frustrating trial and error.

Our engines were the only items off the shelf, so to speak. Kelly agreed with me that if we started from scratch to invent our engines, we would be hopelessly late in delivering the first Blackbird. We chose two Pratt & Whitney J-58 afterburning bypass turbojets, designed in 1956 for a Navy Mach 2 fighter-interceptor that had been canceled before the start of production. But the engine, which would need major modifications for our purposes, had already undergone about seven hundred hours of testing before the government cut off its funding. Each of these engines was Godzilla, producing the total output of the Queen Mary’s four huge turbines, which churned out 160,000 shaft horsepower. Using afterburners at Mach 3, the exhaust-gas temperatures would reach an incredible 3,400 degrees.

This propulsion system would not only be the most powerful air-breathing engine ever devised but also the first ever to fly continuously on its afterburners, using about eight thousand gallons of fuel an hour. To build this system to our needs and specifications, P & W’s chief designer, Bill Brown, who had worked closely with us on the U-2, agreed to construct a separate plant at their Florida manufacturing complex exclusively for developing this extraordinary engine. The CIA unhappily swallowed the enormous development costs of $600 million. Brown preached teamwork and pledged an unprecedented degree of partnership with the Skunk Works in general, and with me and my team in particular, to design their compressor to match my airflow inlet. This close partnership between the engine builder and the airplane manufacturer was unusual in an industry where the engine people and the airplane manufacturers often used each other as scapegoats if an airplane failed to live up to its potential. Abandoning this kind of adversarial posturing led to achieving the most powerful engine system coupled to the highest-performance inlets at these high Mach numbers that has ever been attained.

Bill Brown also offered us access to one of the largest and costliest computer systems of the day, the IBM 710. The system was state-of-the-art for its time and about as sophisticated as today’s commonly used handheld calculators. But, like us, the Pratt & Whitney team would problem-solve mostly by what Kelly jokingly referred to as “my Michigan computer”—the battered old slide rule he had been using since his university days at Michigan.

Despite the unprecedented power of those two massive engines, they supplied only 25 percent of the Blackbird’s thrust at Mach 3, a fact Bill Brown hated to admit. The inlets produced most of the propulsive thrust by supplying the air required by the engine at the highest pressure recovery and with the lowest drag. At supersonic cruising speed, each of our two inlets swallowed 100,000 cubic feet of air per second—the equivalent of two million people inhaling in unison. Hydrocarbon fuels like kerosene burn at high pressure, but at 80,000 feet, the air density is only one-sixteenth the density at sea level, so we used the inlets to pump up compression, before burning the air-fuel mixture inside the engine and then expanding it through a turbine and finally refiring it with tremendous thrust through the afterburner.

The only way to get energy out of the air is to pump pressure into it or to burn it. Our unique movable inlet cone, shaped like a spike, acted as an air throttle by regulating the airflow into the inlet across the spectrum of speeds from takeoff to climb to maximum cruise speed. Operated by our revolutionary electronic measuring sensors, which recorded speed and angle of attack to position the spikes precisely, the movable spikes were fully extended about eight feet out from the inlets on takeoff and gradually retracted by as much as two feet into the inlet interior as the airplane gained maximum supersonic speed.