In the years immediately following World War I, the US Navy’s strong voice in Congress, Admiral William Moffett, became concerned over the unreliability of the liquid-cooled engines powering a brand-new kind of airplane, those deployed from the controversial and experimental aircraft carriers. Fully half of the squawk-sheet complaints and accident reports in Navy files concerned water leaks and other problems with liquid-cooling systems. Because the Navy flies over water, any failure was a serious matter.

Moffett therefore sent two of his young engineering officers, Cdrs. Leighton and Wilson, on a trip around the nation to report on the status of aircraft engine and accessory development. The sense of their report was that air-cooled radial piston engines, if given the same level of development that liquid-cooled engines had already received, would be a far superior powerplant—more reliable, lighter, and faster-climbing. The latter was of prime importance, for fleet defense required development of deck fighters capable of quickly climbing to the altitude of incoming attackers.

The miracle here is that Moffett actually took the expert advice of those lower-ranking men. He let it be known to the industry that the US Navy would buy no more liquid-cooled engines after the expiry of existing contracts. Further, he announced that the Navy intended to buy 200 air-cooled radial engines delivering a minimum of 350 hp from a weight of no more than 650 pounds.

Opportunity knocked: Wright Aeronautical Corporation president Frederick Rentschler had quit that major engine maker over chronic disagreement with the company’s board. They wanted to pay a dividend, but he knew the money was essential to keep R&D strong. When he learned of the Navy’s move, he scurried around raising finance and gathering former Wright engineers to his cause. Because of the name already on the building they would lease—Pratt & Whitney Machine Tool Company—the new operation was called Pratt & Whitney Aircraft.

The first P&W Wasp of nine cylinders and 1,340 cubic inches was finished December 24, 1926. On test, it delivered 405 hp on the first pull. Not only did the air-cooled radial power the great majority of aircraft in WWII, it then went on to power the worldwide development of commercial aviation.

The leading Allied liquid-cooled combat engine of WWII was the Rolls-Royce Merlin, which powered the British Hurricane and Spitfire, then the US Mustang as well. But when it was simplified for use in commercial aviation postwar, that powerplant attracted few orders.

Soichiro Honda strongly believed in the simplicity and directness of air-cooling.

Air-cooling was ideal for aviation use, as the high speeds of aircraft could easily push air through the very closely spaced cooling fins required. Service was simple: Any cylinder on a radial engine could be replaced in a couple of hours. But a piston failure on a water-cooled required lifting a cylinder head and detaching its connecting rod from the crankshaft, a much bigger job. As anyone who watches the Canadian TV series Ice Pilots NWT has seen, big air-cooled radials continue to power DC-3s, C-46s, and DC-4s in Arctic flying, thanks to the huge stocks of parts produced for such engines during the war more than 70 years ago.

Soichiro Honda was a man who believed what was in front of him. Every one of his company’s high-revving four-valve four-stroke roadracers of the 1960s, such as the fabulous RC166 250cc six-cylinder ridden by Mike Hailwood, had been air-cooled. As he and his lieutenants planned the company’s transition to auto manufacturing, Honda strongly believed in the simplicity and directness of air-cooling. After all, he noted, even with water-cooling, the heat is ultimately dissipated to air, so why not eliminate the middleman?

When Honda in 1965 entered Formula 1 auto racing, its first engines were V-12s built very much as the winning bike engines had been, except that they were water-cooled. They revved very high, had built-up all-roller crankshafts and four valves per cylinder. Soon Mr. Honda insisted that an air-cooled 120-degree V-8 F-1 engine should be built, the RA302, to further prove his concept. At the same time, the company designed the H1300, an air-cooled production automobile.

RA302 Formula 1 car
More than a half-century ago, Honda built the RA302 Formula 1 car, which was powered by an air-cooled 120-degree 2,991cc V-8. This example is housed at the Honda Collection Hall in Motegi, Japan.Honda

Mr. Honda had tackled a major objection to air-cooling, that of excessive noise, by treating cooling air just like cooling water by sending it through a coolant jacket surrounding the hot parts. The fins themselves inside this jacket were short and stubby, too stiff to “ring” as conventional fins did (often breaking off from this vibration!). This was Honda’s Duo Dyna Air Cooling, or DDAC. At each point of objection, Honda and his engineers had created solutions.

Project leader on the 1300 was Tadashi Kume, the man who had designed Honda’s first Grand Prix bike engine, the 125cc twin raced at the Isle of Man in 1959. Kume had joined the company in 1954! Another top H1300 project engineer was Kimio Shinmura, who had previously designed the 1960 250cc inline-four RC160, which had evolved into the line of Honda fours that would win many championships and make the Honda name world-famous.

Problems began to pop up for H1300 faster than they could be solved, requiring constant design changes, retooling, and rearrangement of supplier contracts. Soon this work required 24-hour staffing and a special bureaucracy, yet still the problems appeared. In April of 1968, Honda president Kiyoshi Kawashima stopped the production line, then reversed it as previously completed cars were disassembled to receive the changes that had been found necessary in the meantime.

As H1300 gradually struggled into production, another blow fell. When primary driver John Surtees, himself a multi-time 500cc world champion, declined to drive the air-cooled RA302, French driver Jo Schlesser was entered. An unfortunate crash on lap two of the 1968 French GP put an end to Schlesser and to the air-cooled GP car. Honda at the end of that year withdrew from F-1 for a long time.

When, despite praise for its technology, the H1300 car failed to sell in large numbers, in July 1969 Mr. Honda’s longtime business partner, Takeo Fujisawa, called a meeting of 50 engineers from the R&D center. Over drinks afterward, R&D chief and former F-1 project manager Hideo Sugiura told him, “We do not think the air-cooled engines will work.” At a meeting in Atami, Fujisawa authorized a confrontation. Sugiura and his team drove to the R&D center where they told Mr. Honda, “If we don’t change course, we won’t be able to satisfy the emissions control requirements.”

Sugiura later recounted that when Honda replied, “Do whatever you want.” He thought, “That was the moment when I knew Honda was saved.”

After this powerful lesson, engineering quickly altered course, developing the lean-burn Compound Vortex Controlled Combustion (CVCC) process that astounded the US auto industry by meeting early US emissions standards without exhaust post-treatment. A liquid-cooled CVCC engine powered the original Honda Civic, which has been a long success story.

As Mr. Honda liked to say, “You can learn more from failure than from success.”

Liquid-cooling was the correct solution for auto engines because by holding temperature constant across the wide range of auto operation, it simplifies emissions compliance and many other problems.