In February 1956, six months before the U-2 flew its first operational mission, Clarence "Kelly" Johnson's Skunk Works quietly began work on its successor. Project Suntan, officially the Lockheed CL-400, was designed to cruise at Mach 2.5 at 99,000 feet — twice as fast and 30,000 feet lower than the U-2, but far above any Soviet interceptor. The trick? It would burn liquid hydrogen.

The Air Force funded it with $95 million from black budget lines that never appeared in Congressional testimony. By late 1957, Skunk Works had two airframes under construction in Burbank and Pratt & Whitney had a working engine — the Model 304 — that actually ran on LH2 on the test stand. Garrett AiResearch built the pumps. Air Products Inc. was contracted to produce liquid hydrogen at industrial scale for the first time in history, building a plant at West Palm Beach that eventually cranked out several tons per day.

Hydrogen was chosen for one reason: specific energy. Hydrogen has roughly 2.8× the energy per kilogram of JP-4. In principle, an LH2 aircraft can fly farther on less mass. In practice, LH2 has 1/11th the density of jet fuel, so the tanks are enormous. The CL-400's fuselage was essentially two giant vacuum-jacketed cryogenic thermoses with a cockpit strapped on the front, wings on the sides, and two podded engines hanging on stub pylons.

In October 1957, Johnson ran the numbers again and didn't like what he saw. The range, he calculated, would top out around 2,500 miles — barely half the 4,000+ nautical mile mission the Air Force needed to overfly the USSR from friendly bases. In February 1958, Johnson wrote directly to General Donald Putt at the Pentagon recommending cancellation. Within weeks Suntan was dead. The two airframes were scrapped. Lockheed pivoted the same team to what became the A-12 Oxcart — kerosene-burning, titanium, Mach 3.2, and eventually the SR-71.

But Suntan didn't die uselessly. The hydrogen infrastructure Air Products built for it became the industrial base that fed the Centaur upper stage, which flew in 1962 and still puts spacecraft into orbit in 2026 as Centaur V on Vulcan. The Pratt 304 engine data fed the RL10. Every hydrogen-burning rocket engine in service traces some ancestry to a spy plane that never flew.

Why it's worth revisiting now: Johnson's range problem was fundamentally a tank problem — 1957 metallic tanks were heavy and leaked. In 2026 we have composite cryogenic tanks (Boeing's X-33 tank, Blue Origin's New Glenn LH2 tank), aerogel and multi-layer vacuum insulation that cut boil-off by an order of magnitude, and industrial hydrogen production at scales Air Products in 1957 couldn't have imagined. Airbus ZEROe, ZeroAvia, and Universal Hydrogen are all building LH2 aircraft targeting entry into service before 2035. NASA's 2010 X-43 hydrogen scramjet work resurrected some of the aerothermal data Suntan generated.

A modern CL-400 — a high-altitude, long-endurance hydrogen platform for persistent ISR, atmospheric science, or communications relay — would carry perhaps 40% more fuel in the same tank volume as 1957 hardware, with vastly reduced boil-off. Johnson's 2,500-mile shortfall was a materials problem, not a physics problem. The physics still works. The tanks finally do too.