A recent video shows Sammy Miller, veteran British rider and now museum curator, discussing a cylinder from a small aircraft radial engine with Allen Millyard, who builds whatever his imagination proposes. For example, a V8 using two four-cylinder in-line Yamaha top-ends on a common crankcase.

As they speak, Miller casually observes that the cylinder’s fin pitch is closer than was used on air-cooled motorcycle engines. Finer pitch cooling fins were tried on motorcycle engines, but they raised operating temperature rather than lowered it as hoped.

The reason for this is the presence, near fin surfaces, of a growing layer of stagnant air that has been slowed by frequent collisions with the metal surface. The farther the air flows, the thicker this “boundary layer” becomes. In narrow fin spaces this increasingly obstructs airflow through them, causing air to flow around the cylinder rather than through its fin spaces.

Many aircraft in WW I were powered by rotary air-cooled radial engines—meaning that the crankshaft was stationary, bolted to the firewall, while the crankcase, cylinders and pistons rotated around it. Rotating the whole engine provided cooling that the low air speeds of 1914 could not. The big problem was that while the leading faces of the cylinders cooled fairly well, their trailing faces received almost no cooling. They overheated and bulged out-of-round such that no normal piston ring could seal them. A brass L-section ring resembling the leather seal washer in an old bicycle pump was used instead – its flexible vertical part was pushed out against whatever the cylinder wall was doing by gas pressure. Such “obturators” needed replacement every 10-24 hours.

Because motorcycles seldom cruise at very high speed, their cylinders can be designed with wider fin pitch of .300 - .400”, backed up by a “heat sink” of extra metal to absorb much of the extra heat generated by bursts of high speed. That heat is then carried away during the next off-throttle period.

Aircraft piston engines used maximum power for take-off, then made a power reduction to climb power, and once at altitude cruised at a constant percentage of max power – usually 50% or a bit more. A heat sink is of no use here, so the aircraft cylinder’s heat rejection to cooling air must keep pace with engine power level (or the engine will gradually get hotter and hotter until it fails).

What Mr. Miller noticed was that aircraft engines used a different scheme for cooling than motorcycle engines. That was pressure cooling. Yes, the presence of boundary layers opposed airflow through the fin spaces, but given enough pressure difference across those spaces, it was possible to achieve rates of airflow through them that would maintain desired safe cylinder head temperature during steady operation.

Open any book on the history of aviation, or walk into any substantial air museum and you will see the evolution of aircraft engine cooling.

Early radials operated with their cylinders and heads just sticking out into the airflow. As R&D made engines ever-more powerful (The airlines wanted power, the Air Corps wanted power!) this kind of cooling became inadequate. At the same time, the drag of those jutting cylinders was tremendous, cutting top speed.

Now those round engines were enclosed in streamlined nacelles, and triangular baffles appeared between cylinders, making it harder for air to flow around them. Gradually, the cylinder fins were enclosed by sheet metal baffles, open at the front and rear, shaped so as to guide airflow all the way through the fins rather than around them.

This process – strongly resisted by Wright, who believed a gap must exist between baffles and fins to prevent vibration from breaking them – took ten years. At its end, the only flow path through the engine nacelle was through fin space. A bee, entering the nacelle at the front, would be unable to pass through. In historical accounts, this tight cowling is often blamed by uneducated persons for overheating, but the facts are otherwise. At the end of the piston-engine era, Wright’s big post-war R3350 (sharing zero parts with the wartime 3350) was in turbo-compound form making 3700 hp for take-off, while P&W’s monster 4-row 28-cylinder R-4360-63A was making 3800. These high powers would have been impossible to cool without tight pressure baffling.

When I mentioned the growth of engine power, I wasn’t speaking of something subtle. Wright’s great single-row 9-cylinder R-1820 began service at just over 500 hp, and at the end of its long career it was giving 1525 hp.

To make this possible, every increase in power had to be met with closer-spaced cooling fins of ever-greater depth. Even then, some improvisation was necessary to make things work as planned. When the Wright R-3350 engines on the B-29 were first in service, it was predicted that an airspeed of 140 mph would push airflow through the fin space at a rate high enough to maintain safe cylinder head temps (CHT) during climb. When this proved to be not enough, pilots were instructed not to climb immediately after take-off from their island bases in the Pacific (Tinian, Guam, and Saipan) but cruise at 200 mph not far above the water for some time to reduce fuel load, eventually making it easier to make the arduous climb to bombing altitude—30,000 ft. When it was shown that the low-density air at that altitude could not adequately control CHT there was big discussion.

High CHTs around 500F were above the 40% of melting-point temperature that is the threshold of the process of metal creep. Hot metal under stress can slowly, slowly yield. Operating and residual stresses at this temperature were allowing exhaust valve seats to ovalize (two points of contact instead of full, cooling contact with the valve seat) or to move off the centerline of valve and guide (only one point of seat contact).
Without the cooling effect of full seat contact, valves were leaking hot exhaust gas, further raising their temperature. Eventually valve heads were broken off by the piston when their stems seized in their guides. Velocette racing manager Harold Willis in the 1930s called a valve head, hammered into a piston crown “penny-in-the-slot". This could result in jacking an entire cylinder off the forged steel crankcase, often causing an engine fire.

There began to be discussion of providing cooling blowers, but a quicker solution was already “in the air.” Unit commanders were calling for night operation, with bombing altitudes reduced to 5000-8000 ft. This worked, enabling B-29s to burn out Japan’s 66 largest cities, beginning the night of March 9/10, 1945.

The six-engined B-36 bomber, originally conceived as a means of bombing Germany had England been forced to surrender early in WW II, was designed with engine-driven cooling blowers to enable normal engine operation at 40,000’. They consumed roughly 10% of engine power. Another fan-cooled aircraft was Germany’s FW190 fighter, powered by BMW’s excellent 801 14-cylinder radial.

When plans to further up-rate the B-36’s big piston engines failed, the quick-fix of the moment was to hang a few jet engines under the vast wing “to help out.”

Fan cooling can be highly effective – Porsche has shown what can be achieved with it. But a cooling fan on a motorcycle? Who could survive the drastic loss of face associated with rude skeptics asking, “What is that - a golf-cart motor?”

The arrival and fast development of turbojet and turboprop engines put a stop to such nonsense. Only about ¼ of the air they inhaled took part in combustion. The rest was used for a variety of internal cooling duties. At take-off, each blade on an airliner’s HP turbine disc is encased in a flow of cool air emerging from laser-drilled orifices in the hollow blade. The gas flowing past these blades is at a temperature greater than their melting-point.