Pratt and Whitney's R-2800 cu in. (45.9 L) was America's first 18 cylinder radial, the Double Wasp. Much smaller than the world's only other modern eighteen, the Gnome-Rhone 18L of 3,442 cu in. (56.4 L), it was nevertheless more powerful, and heat dissipation was correspondingly more of a problem. This meant that, for the R-2800, the cast or forged cooling fins that had served so well in the past had to be discarded. The cooling fins needed were so thin and fine-pitched that they had to be machined from the solid metal of the head forging. All the fins were cut together. A gang of milling saws was automatically guided as it fed across the head so that the bottom of the grooves rose and fell to make the roots of the fins follow the contour of the head. The results were worth the trouble as it was a case of designing an engine component that could only be made by a new method and then keeping everything crossed until the new method proved to be practical. In addition to the new head design, the Double Wasp had probably the most scientific baffling yet to direct the flow of cooling air, more so even than the excellent arrangements on the Ranger inline air-cooled engines.

2,000 hp was obtained from the R-2800 with 1 hp/1.4 cu in. (43.6 hp/L) of displacement. In 1939, when the R-2800 was introduced, no other aircooled engine came close to this figure, and even liquid-cooled ones barely matched it. The designing of conventional air-cooled radial engines had become so scientific and systematic by 1939 that the Double Wasp was introduced at a power rating that was not amenable to anything like the developmental power increases that had been common with earlier engines. It went to 2,100 hp in 1941 and to 2,400 late in the war, but that was all for production models. Experimental models, as always, were coaxed into giving more power, one fan-cooled subtype producing 2,800 hp, but in general the R-2800 was a rather fully developed powerplant right from the beginning.

It was exclusively a powerplant for fighters and medium bombers during the war, being used in the P-47 , the F6F Hellcat, and the F4U Corsair , and also in the B-26 and A-26 twin engine mediums. Post-war its reliability commended its use for long-range patrol planes and for the DC-6, Martin 404, and Convair transports. This last application is noteworthy, since these were twin-engine craft of size, passenger capacity, and high wing loading comparable with the DC-4 and the first Constellations.

Two engines were all right for transports with the DC-3's moderate wing loading, and the high wing loading of the DC-4 was safe enough when there were four engines, but all that weight with only two engines seemed like tempting fate. However, the Convair engineers knew what they were doing. (Those at Martin, and those who tested the Martin for government approval, didn't; the Martin's wings failed from fatigue after a while.) The Convairs were just as good in their way as the four-engine transports. A well engineered installation and good controls were probably what made the difference.

When the USA went to war in December 1941 there were very quickly some major changes in philosophy. Such long-established engines as the Cyclone and Double Wasp were re-rated on fuel of much higher anti-knock value to give considerably more power. Perhaps the most outstanding example was the great R-2800 Double Wasp, which went into production in 1940 for the B-26 Marauder at 1,850 hp and by 1944 was in service in late model P-47 Thunderbolts (and other aircraft) at a rating of 2,800 (experimental) hp on 115-grade fuel with water injection. Of course, all engines naturally grow in power with development, but a major war demands the utmost performance from engines fitted to aircraft whose life in front-line service was unlikely to exceed 50 hours' flying, over a period of only a month or two.

In peace time the call was for reliability over a period of perhaps a dozen years. And of course a pilot in combat has no time to fiddle endlessly with a fistful of engine controls in order to maintain the optimum engine operating conditions, and bearing in mind the rate at which aircrew had to be produced in wartime he probably did not have the knowledge of how to do this either.