Rotary engines

by Martin Gregorie

Rotary engines were commonly used to power single engined aircraft between 1913 and 1920. During the 1920s they became obsolete as other types, especially radials, achieved better power to weight ratios and were easier to manage during flight.


Rotary engines were built 'backward'. Where a normal engine has its crank case bolted to the aeroplane and the prop attached to its crankshaft, the rotary had its crankshaft attached to the aeroplane and the propeller bolted to the crank case. The whole engine spun with the propeller. Rotaries were much lighter for their power output than other types of aero engine and could turn a very large propeller at fairly slow speeds: 1300 rpm was a typical maximum engine speed. This was beneficial because a large, slow turning prop is very efficient. By contrast most large inline and radial engines had to be geared down to match the engine to the most efficient propeller and this required a heavy gear box. An additional benefit was that rotaries are almost vibration-free in operation: they are the smoothest of all piston engines. The consequence of fixing the crankshaft to the airframe is that the pistons do not move in relation to the airframe while the rest of the engine just rotates. As a result there are no oscillating motions within the engine and so very little vibration is generated.

The biggest drawback of the rotary was a direct consequence of it's rotation. The engine weighed typically 320 to 340 lbs, or something like 25% to 33% of the entire airframe weight. This sort of mass rotating at over 1000 rpm makes a large and effective gyroscope. All was well in straight and level flight, but as soon as the pilot attempted to alter his aeroplane's attitude or turn it the gyroscopic forces became very apparent, forcing the nose down in a left hand turn and up if he turned right. In consequence a rotary powered plane could turn and dive extremely quickly to the left while a sudden turn to the right was likely to cause a stall and spin. With the rotating engine continuing to hold the nose high in a right hand spin recovery would be very difficult. Jumping out was not an option: the Royal Flying Corps didn't wear parachutes because their senior officers thought pilots would be forced to fight harder if they had no way to abandon the aircraft. It seems not to have occurred to these gentry that a pilot might have a genuine need to exit a disabled or burning machine.

There was a single block-tube carburetor attached to the rear end of the hollow crank-shaft. Lubrication was a total loss system. Castor oil was injected into the carburetor by an engine-driven pump, with the unburnt residue forming part of the exhaust. The fuel/oil/air mixture flowed through the crank shaft and into the crank case, where it lubricated the bearings before being transferred into the cylinders.

Unlike a modern fuel lubricated two stroke, the crank case volume in a rotary does not vary as the engine rotates. As a consequence there is no pumping action to force fresh mixture into the cylinders. Instead, cylinder was filled due to the suction generated inside it as the piston travelled down after the exhaust stroke. As a result the power output per swept volume was low and so rotaries had an enormous swept capacity for their output. The volumetric efficiency was terrible, under 10 hp per liter, but they could be built extremely light and, after all, swept volume weighs virtually nothing! However, if you build a large capacity, low volumetric efficiency engine nice and light the first two factors cancel out and the engine will still have a good power/weight ratio.

Because the only forces moving fuel/air mixture through the engine were centrifugal force in the transfer passages plus suction generated in the cylinders it is likely that the carburetor itself didn't generate enough vacuum to suck petrol through the jet and atomise it, so rotaries used a pressurised fuel system to spray petrol into the carburetor. Pressure was built up before starting by using a hand pump in the cockpit. Once the plane was flying a small propeller-driven air pump kept the system pressurised. A careful look at photos of rotary-engined aircraft will show a propeller and pump attached to a wing strut near the fuselage where it can be spun by the slip stream.

No exhaust system was fitted: indeed it would have been impossible to fit anything like that to a spinning engine. In consequence the exhaust valve on each cylinder opened directly into free air inside the cowling. Normally an engine without an exhaust system is extremely noisy, with a loud bang (the bark) each time an exhaust port opens. This is because modern engine design requires the exhaust valve to open while there is still a lot of pressure in the cylinder. However, rotaries were very low compression and ran slowly. As a result they were relatively quiet in operation despite having no exhaust system because there is relatively little pressure in the cylinder by the time the exhaust opens.

The sound a rotary makes is a cross between a low frequency buzz and a compressed-air hiss. I've heard both scale model rotaries and the real thing running and they all have this unusual sound.

They were 4 stroke engines. The operating cycle was:

Cycle Stroke Piston movement Action
1 1 downward suck in new mixture
1 2 upward compression
2 3 downward power stroke
2 4 upward exhaust

Assuming that the exhaust valve opens at bottom dead centre and stays open until top dead centre certainly explains the lack of any exhaust bark or crackle.

Engine controls

Rotaries were difficult engines to manage. Gross power control was handled via a 'blip switch' on the top of the control stick that switched the ignition on and off. Some types switched the ignition on selected cylinders, but on most engines the switch controlled all cylinders: the engine was either on or off and the aircraft either ran at full power or glided. It was possible to get reduced power settings for landing or to make the engine idle by pulsing the 'blip switch' slowly to give an average output somewhere between an idle and full power. The sound of a rotary flying round the airfield is quite distinctive. As the pilot uses the switch to maintain slow, level flight you hear a BZZZZZZZZZZT............BZZZZZZZZZZT............ noise. A side effect of this power control was that when the aeroplane glided down, coming in to land for instance, oil and fuel was still fed into all the cylinders including those which were not firing. If the engine was off for too long it could be very slow to pick up again. Often as much as 30 seconds was needed to regain full power. This unused fuel was spewed out of the exhaust valves so that it collected in the cowl and on the fuselage as well as oiling up the spark plugs. Meanwhile the pilot got a major dose of castor oil. Switch on again... whoof!!! With any luck, the conflagration would be short-lived. Just as well, because World War I pilots did not wear parachutes.

Another way to incinerate oneself was via an engine back-fire, which was known to start a carburetor fire on occasion. This gave the pilot a case of hot-foot because there was often nothing between his feet and the carburetor.

Although the 'blip switch' was a simple, direct and light weight engine control and adequate for airfield and circuit operation it was tiring to use for formation flying or flying at cruise power for any length of time. None the less, some rotaries just made do with this switch and a fuel lever but others added a power lever in an attempt to make the engine more tractable.

The fuel lever controlled the amount of fuel entering the carburetor: it adjusted the strength of the fuel/air mixture. This is an important control because the fuel flow must be adjusted to suit the day's temperature and humidity as well as the altitude at which the aircraft is flying. An engine will only run if the mixture strength is within a suitable range: if the mixture is too lean or too rich the engine will stop.

Engines that also had a power lever allowed the pilot to adjust the air flow into the carburetor, and hence the power output of the engine when the 'blip switch' was held closed. This made formation flying somewhat easier as well as easing the workload during a reduced power cruise. The range of adjustment was fairly small. Typically the minimum setting reduced the engine speed to just over half the full power value. This was not sufficient to cater for idling or landing approaches, so the 'blip switch' was still needed. The power lever was best thought of as a means of fine tuning the engine's output during normal flight. In any case, for reasons given below, it was not a good idea to fiddle with it when close to the ground.

Each time the power lever was moved the fuel lever needed to be adjusted immediately to keep the fuel/air mixture within operating limits. There was no linkage between these two controls and they did not have a similar effect if moved by the same amount. So, if the pilot needed to alter the power setting he needed to make an immediate and correct change to the fuel setting. If he got it wrong, the engine stopped running though it would still be spinning in the slip stream.

If the power setting had been increased the engine would suffer a 'lean cut' because there was now too much air and not enough fuel entering the engine. The pilot recovered by closing the fuel lever and then slowly re-opening it until the engine restarted. This usually left the plane without power for 5 seconds.

However, if the power lever setting had been reduced the engine would suffer a 'rich cut' because there was too much fuel in the air stream for the mixture to burn. This time the recovery was to immediately close the fuel lever and wait while the engine blew the excess fuel out its exhaust. Once it had dried out the fuel lever was slowly opened until the engine restarted. This process took a minimum of 25 seconds. Hence it not being a good idea to fiddle with the power lever near the ground. If, at low altitude, the pilot suffered a rich cut or, worse, had a rich cut but tried to re-open the fuel lever too soon then he would suddenly find he needed to make an immediate landing without the benefit of an engine. There was also the possibility of an engine fire if there was still a lot of unburnt fuel inside the cowling when the engine started up again.