~ Auto Buzz ~: ASK KEVIN: Why Do Air-Cooled Engines Have More Trouble with Emissions Regulations Than Liquid-Cooled Ones?

QUESTION: Why do air-cooled motors generate more emissions than water-cooled ones? Why can’t they just engineer cooler and cleaner air-cooled engines to meet the new emissions regulations? Silas B. Laubmann Atlanta, GA ANSWER: Differences between air- and water-cooled engines have been with us a long time. Compare two combat aircraft engines of World War 2, making approximately the same power. The air-cooled 14-cylinder Wright R-2600 made 1700 hp in early versions, from approximately 2600 cubic inches (.65 hp per cubic inch), while the liquid-cooled V-12 Rolls-Royce “Merlin” made about the same power from 1650 cubic inches (approximately 1.0 hp per cubic inch), or 50 percent more power per cubic inch. The air-cooled Wright engine was unable because of its higher operating temperature to accept as much supercharger boost as could the Merlin, but the engines weighed a similar amount.

Air-cooled 14-cylinder Wright R-2600.

Liquid-cooled V-12 Rolls-Royce “Merlin” engine.

Now compare a current liquid-cooled MotoGP engine of one liter displacement, making maybe 245 hp @ 16,000, with Rob Muzzy's best air-cooled Kawasaki Z1-based 1025cc Superbike engine of 1982, making 150 hp @ 10,250 rpm. The MotoGP engine, because it is well cooled, can operate knock-free at torque-boosting compression ratios above 13:1. Knock, or detonation, is a heat-driven process; everything that heats the fuel-air charge pushes it toward eventually detonating. This includes the temperature of the intake port and valve, the combustion chamber surface temperature, and the piston crown temperature. The air-cooled Superbike had to run with quite a lot less compression—around 10.0:1—to finish races. At 10,250 rpm, it was pushing its mechanical limits, for its roller crank was designed for a maximum of 7,500 rpm and crankcase flexure required Muzzy to use a solid copper cylinder base gasket to prevent the flex from rolling the production gasket up into little balls. This is a comparison of a 2015 engine with a 1982 engine, so some of the rpm and cooling difference could be made up by designing a new air-cooled powerplant from scratch. But one of the air-cooled features so loved by some is their visible cooling fins. Aircraft engine such as the R-2600 had their fins encased in sheet-metal shell baffles that forced cooling air to flow through fin space all the way around each cylinder to exit from a narrow vertical slot at the rear. This was necessary to (1) reduce drag by making every pound of cooling air carry away maximum heat, and (2) by more nearly equalizing cooling between front and rear of each cylinder to keep it round enough for piston rings to seal acceptably well, and (3) air guidance of this type was necessary to keep critical areas such as exhaust valve seats from overheating and distorting such that they could no longer seal. This leakage quickly overheated valves, causing them to warp, gutter, or break (there were many such failures).

Would fanciers of air-cooling be willing to accept shell baffles of this kind? The reason why Harley, BMW, and soon, Triumph, are resorting to localized liquid cooling is because they, too, have encountered cylinder head distortion, valve leakage, and valve problems. As to the emissions problem, a cylinder at constant temperature from liquid cooling can be relied upon to heat the air entering it a predictable amount. But an air-cooled cylinder and head run hot in summer, cool in winter. To give an air-cooled engine some ability to protect itself from short periods of high-power operation, its head is made quite heavy to act as a heat sink. This makes it take a while to come to operating temperature. Certainly such differences could be written into software and temperature sensors in the head could report the necessary information. Cylinder distortion and the extra clearance necessary to deal with hot and cold add up to some issues in oil consumption. Pistons in air-cooled engines must have their rings located farther down to keep them away from high temperature that gums oil and sticks piston rings. This creates more emissions from the unburned charge that is forced by compression and combustion into the ring crevice volume on each cycle.

The original engines from which the present Harley-Davidson and BMW powerplants have been developed had smaller displacement. As an engine is made larger, what becomes important to exhaust valve temperature is the volume of hot gas flowing through it. But the exterior engine surface on which cooling fins can be made increases less rapidly than does displacement (area is dimension, squared, while volume is dimension, cubed). As the makers of these older classic designs have made their engines bigger to keep up with owner expectation (to ride up the Rockies in summer, with wife on the pillion seat, and towing a loaded vacation trailer, at the 75 to 85 mph at which traffic moves now), engine exhaust volume has grown faster than cooling fin area, leading to the need for concentrated cooling. Water is 830 times denser than air, so that defines the problem of air-cooling. The air-cooled BMW 801 radial in the FW-190 fighter plane was given a cooling fan (which I have found motorcyclists generally hate, considering it suitable only for scooters). Motorcycle engines used to be universally air-cooled because that was the easiest solution in those times. Today, it has become much more difficult than liquid-cooled. Surely a lot of expensive R&D could produce workable, emissions-legal air-cooled motorcycle engines—Honda engineers have described such a process in a story on the Internet. That would imply a price premium. Would the market cheerfully accept it? Send your “Ask Kevin” questions to cwservice@cycleworld.com. We cannot guarantee a reply to every inquiry.