Why Liquid Cooling For Motorcycles? EPA compliance is only part of the reason


Motorcycle engines have had cooling fins since the beginning, so it was a shock for many riders when the arrival of liquid cooling began to make them unnecessary. One response to this has been to cast decorative nostalgia fins onto the outsides of engine water jackets.

I suspect that the liquid cooling of Honda's original GL1000, unveiled in late 1974, owed much to that company's development of the low-emissions 'Civic' auto, with its lean-burn 'CVCC' combustion system. Many Honda engineers were involved in that work, which had come hard on the heels of the "Revolt of the Engineers" in 1968. At that time, development of Mr. Honda's H1300 air cooled auto ran into so many problems that his long-time business partner Takeo Fujisawa sided with the engineering staff in supporting a shift to liquid cooling. The new focus on the automotive side was emissions control, which had become a major priority in the US market.

Air cooling made emissions control more difficult because the temperature of such engines varies with the weather¡ªhotter in summer, colder in winter. As the fuel-air mixture enters an engine cylinder on its intake stroke, that mixture is heated to the temperature of the surfaces surrounding it¡ªthe cylinder wall, head, and piston crown. This is fine if engine temperature is constant¡ªwe just jet the carburetor to deliver the desired mixture, knowing that air density in the engine is constant.


Air cooling makes emission control more difficult because of the varying climate in which a motorcycle is operated.Courtesy of Royal Enfield

But an air cooled engine¡¯s temperature constantly rises and falls with weather changes and engine power output. Air cooled engines become extremely hot at high power levels, but cool off comfortably at idle and in leisurely riding. The hotter the engine becomes, the more the fuel-air mixture entering it expands and loses density. This not only loses power in proportion to the mixture density lost, it also loses power because, with the air less dense, the mixture becomes richer.

This is further complicated if the rider is a die-hard (as I was in the winter of 1968) who rides year round. Low winter temperature increases air density, so there is more air in proportion to fuel¡ªa lean mixture condition. If the carburetor is set to deliver a correct mixture in winter, it will be very rich in August.

Racers had no problem with this¡ªthey were used to rejetting their carburetor(s) several times a day to maximize power and response. But riders of production bikes just want to go riding, so their carburetors were set to a compromise¡ªlean in winter, rich in summer.

Then came the EPA and unending pressure to reduce engine exhaust emissions. Given the limits of carburetor fuel systems, the quickest way to making fuel-air mixtures more constant year-round was to make engine temperature constant by adopting liquid cooling regulated by thermostat. So that is just what the motorcycle industry did during the 1980s¡ªmostly.

Okay, you may object, I get it. But today carburetors are gone, replaced by closed-loop mixture control via digital fuel injection and an oxygen sensor in the exhaust. Surely that system can cope with variations in engine temperature? Yes, DFI can deliver a constant mixture, but it can¡¯t recover the horsepower that is lost to reduced air density when an air cooled engine becomes very hot.

Engine oil to span the wider temperature variation of an air cooled engine must either be a wide-range multi-grade, or different oil viscosities must be specified for summer and winter. A wide-range multi-grade in an air cooled engine suffers rapid evaporation of its light base oil (for example, the 10W in 10W-40) when its cylinder walls are very hot in summer, adding unburned hydrocarbons (UHC) to the exhaust stream or pushing them out the crankcase breather. Using different oil viscosities winter and summer goes against the modern trend toward minimum maintenance.

Another problem is temperature-driven changes in engine clearances. The crankshaft is steel but the aluminum crankcase that carries it is made of aluminum, which expands with heat three times more than steel. So in summer, when the oil is thinnest, bearing clearances are at their largest. F1 engines run synthetic oils so watery that the tiny bearing clearances they require won¡¯t allow the starter to turn the engine until it¡¯s been pre-heated by circulating hot coolant through it. Thermostatic liquid cooling means clearances remain constant.

Pistons run hotter in air cooled engines because all they have to cool them is contact with the quite hot cylinder walls. Therefore such engines tend to employ longer-skirted and heavier pistons than the light "ash-tray" pistons found in liquid cooled designs (which usually have piston-cooling oil jets). That extra piston material is there to act as a "heat pipe" to conduct heat from the piston crown to a wide area of contact with the cylinder wall. Heavier pistons equate to increased vibration and bearing loads, but they were the norm 40 years ago.

Next, let¡¯s try to keep piston clearance small and constant across the air cooled engine¡¯s wider temperature operating range. Not so easy, and all sorts of weird effects turn up when pistons tilt and clack from thrust to non-thrust face in the presence of larger clearance. Piston ring motion can act like a miniature oil pump, scraping oil off the cylinder wall only to have it end up mixing into combustion air, where it becomes UHC out the exhaust valve(s). Never mind, we¡¯ll just tell engineering to keep trying stuff until piston clearance is stabilized under all conditions. More or less.

liquid cooled twin cylinder

A liquid cooled twin cylinder from Kawasaki¡¯s Ninja 400. It¡¯s easier to achieve low emissions with a liquid cooled engine, and the number of air cooled models become less and less.Courtesy of Kawasaki

Then there¡¯s the problem of compression ratio. Ever notice that liquid cooled motorbike engines have high compressions up in the 12¨C13:1 range, while the classic air coolers of the 1970s seldom had compression above 9.5:1? Even modern air coolers are usually below 10.5:1. The reason air cooled engines must use lower compression is that when operating at the high end of their temperature range, any higher compression would cause them to detonate (detonation¡ªaka ¡°engine knock¡±¡ªis an abnormal and often destructive form of combustion that becomes more likely as mixture temperature rises). In general, the higher an engine¡¯s compression can safely be set, the greater the torque it delivers and the lower its fuel consumption.

Anyway, you get the general idea¡ªit¡¯s easier to develop a successful liquid cooled engine of good performance and low emissions than it is an air cooled¡ªno matter how much we may have loved those cooling fins.