Sam, You Made The Fins Too Short!
One morning this week I awoke filled with resolve to know more about air cooling -- mainly driven by continuing curiosity about the problems of the engines used on the B29. While one of my sons fried hashbrowns and bacon downstairs, I pored over the excellent aerodynamic cookbook, "Fluid Dynamic Drag", by Hoerner. It derives a nice, simple relationship that says that drag horsepower -- the power consumed in flowing air through whatever cooling surface arrangement you might have -- is directly proportional to the total cross-sectional area of cooling flow path, and proportional to the 3/2 power of the pressure drop across this flow path.
This says that it will cost less power if we increase the cooling flow path area to get increaed cooling, than it would if we increased the pressure drop across the fin/baffle arrangement by some such scheme as using a blower to push more air through, &c. The historical result of this little equation is to be seen the obvious increase in "solidity" of air-cooled radial engines from the days of the J5 Whirlwind that powered 'Lucky Lindy's" transAtlantic flight to the later R975s of the WW 2 era; cylinders became wider as fins were made deeper to obtain the increased flow path area.
Sometimes there isn't time or room or whatever, to permit such increase. One such case is the BMW 801 engine used on Germany's wonderfully-packaged FW190 fighter. The quickest way to increase engine cooling was to cool it with a blower, and this is why the comparatively beloved B17 aircraft's R1820 engines did cool). But in the war, such a switch involved too much pushing and shoving, and instead, big flaps were riveted to the propeller blades in front of the nacelles' cooling entries, to shove more air through. The B29 had so much pioneering in its design that during climb with the original engines, the pilot and flight engineer had to choose between struggling along at low altitude and speed with the recommended cylinder head temperatures, and with the cowl flaps open so far that the aircraft couldn't accelerate - or close down the cowl flaps, let head temps soar to really frightening values, and have the machine climb laboriously and dangerously to altitude, with the concomitant danger of the aforementioned in-flight fires. Early-model B29s and engines were so closely-designed that cooling drag made all the difference between climbing and not climbing.
In one of my ancient textbooks there is an example that is relevant here. It asks the student to calculate how much more power a given engine could make if something (most likely fuel of a higher octane rating) made it possible to raise cylinder head temperature from 325 deg. F to 355 deg, without detonation. The answer is 25%. In the case of the B29, even with the highest-available fuel octane, supercharging the engine to the point of near-detonation didn't make enough power to do the job.
Later, these Wright R3350 engines went on to be developed into highly reliable powerplants both for the military and for the many Constellations and DC7s they powered -- with forged heads, fuel injection, and other sundry improvements.