The Wankel system was invented by the german engineer Felix Wankel, and first used for pumps and compressors. Only later was the principle used for internal combustion engines. The german car manufacturer NSU produced some models that featured Wankel engines. Later in history, Mazda produced a famous sports car based on a Wankel engine. The Wankel was also used in motorcycles, generator units, airplanes, boats, as an auxiliary engine for sailplanes, and even a very large Wankel for use in tanks was developed.
The use of Wankel engines in daily life has nowadays become quite limited, but the design survives in non-engine form: Many of today's vacuum pumps still are based on a Wankel design. And for model fliers, there is a jewel available: In the 1970s the japanese model engine specialist O.S. Max, and the german model building company Graupner, jointly developed a model Wankel engine. I was lucky enough to get one of these marvels, in a way I had never expected: I saw it in a stationery shop that had a little appendage of model building material. The price tag was written on the box. I looked twice. It couldn't be possible! At least one zero had to be missing! The price stated there was cheap for a standard 2-stroke engine, impossibly cheap for this special, unique thing! I asked. The store owner confirmed the price, and told me it was a second hand engine, never used, but stored for a long time. It should be OK, he added.
I tried to crank it over. No way. It was stuck like if it was cast from
one piece... That's no surprise with an engine that has been stored for
years, but it prevented any kind of test at the shop. I hesitated for a
minute, then I made up my mind: Even if the engine should be unuseable,
it was worth that little money just for the fun of owning a real Wankel!
I bought it. Or should I rather say, I adopted it.
Once at home, I started disassembling it. I had to clean the rockyfied oil out, and by the way I wanted to inspect its internal condition. First I removed the mount, the silencer and the carburetor, which by the way is a very simple barrel type, without even an air bleeder hole! This engine doesn't need any such tricks for a good idle, I found later...
This picture shows the engine core, stripped of all external parts,
ready for serious disassembly. By this time, despite an experiment done
by heating it, there was still no way to crank it over. I was prepared
for the worst, like broken parts fused together. But as I proceeded in
the disassembly, it became clear that this engine had never even seen any
fuel, and the only reason for it being stuck was the hardened oil!
The top cover is off, revealing the Wankel's inner life. This is a good picture for a technical discussion. The triangular rotor with the rounded sides (a Releaux triangle) has its center mounted by means of a needle bearing to the crank on the crankshaft. The underside of the rotor has an inner gear, that engages with a three times smaller fixed gear on the backplate. The result is that for each revolution of the crankshaft, the rotor describes one orbit, coupled with one third of a revolution. All clear?
The tips of the triangle slide along the walls of the chamber. Spring loaded apex seals keep everything airtight. Thus three chambers are formed, each of which works in a 4-cycle process, with 3 of the 4 processes happening simultaneously at any given time. The 3 chambers expand, contract, and travel around the inside of the engine.
The rotation is leftwards in this engine, so the port on the upper right
is for admission, the one at the lower right is exhaust, and the glow plug
sits in a recess at the middle of the left side. This photo pictures the
instant when the upper right chamber would be in the midst of the admission
cycle, the chamber at the left would be finishing compression and be very
close to ignition, while the lower chamber would just be starting its exhaust
cycle. A moment later, the crankshaft advanced a half turn or so, the rotor
would have rotated by about 60 degrees or so, had a half orbit, and the
upper chamber would be finishing admission and starting compression, the
left lower chamber would be in the midst of the combustion cycle, while
the lower right chamber would be close to finishing exhaust. Now all is
clear, I hope? :-)
It's hard to imagine this process, but having the engine in the hand and playing with it is very enlightening, and makes it easy to understand. I just wonder how in the world Mr. Felix Wankel could invent this crazy thing?!
One of the greatest advantages of the Wankel engine is that it achieves 100% mass balance. Piston engines have both turning and swinging parts, and the swinging ones can be mass-compensated only in multicylinder engines. Model piston engines are mostly single-cylinder designs, thus there is no way to implement full mass compensation, and the engines produce very strong vibrations. Not so with the Wankel! The rotor is balanced in itself, and its orbiting around the crank is purely circular, so it can be perfectly compensated by two counterweights. This photo shows one of them, mounted in the back of the engine, while the other one is integrated in the propeller mount on the front side. The result is perfect mass compensation, and an engine that runs almost without any vibration! The only vibration remaining is in torque-mode, and comes from the combustion cycles, but even this vibration is much lighter than that of a piston engine. The Wankel is almost as smooth-running as an electric motor!!!
Here you can see my Wankel soaking in a bath of lacquer thinner mixed with alcohol. This magic recipe removed all hardened oil, and left my engine shining like new! I dried the parts, applied a light oil coating and reassembled it. Now it turned easily. Very easily, indeed. A little bit too easily! You would expect some compression in an internal combustion engine, don't you? Well, this engine had almost no compression at all! I thought this may be normal, after all it's a fast turning machine. I mounted it on a test bench, fueled it up, fired up the glow plug, and cranked it over until my starter battery was empty. No way. Not even a single cough. Nothing. This engine was dead as a doornail! It was clear enough, with alcohol fuel and a glow plug, without compression there is no ignition, period! Back to the workshop...
Look closely at this picture. The first thing that is obvious is the very coarse surface finish of the rotor. And this surface is supposed to seal tightly against the front plate! When I saw it for the first time, I guessed this roughness may be intentional, to retain some oil or so, but this theory now quickly vanished.
Look very closely at the area at the apex seal. You can see that the seal is longer (standing higher) than the rotor! Time to measure...
The rotor turned out to be 22 microns thinner than the external case. The apex seals had just 5 microns total clearance. 22 microns! Imagine what happens under the pressures that exist during combustion, when the oil is as thin as water in the hot engine! No wonder there was no hint of compression!
The repair work took some dedication. First I sanded the groovy surfaces of the rotor, until they were as smooth as mirrors. For this job I used sandpaper on a highly flat surface, lubricating with oil, and rubbed the rotor in circular motions over it. Every few strokes I checked the dimensions, striving to obtain as good a flatness as possible, while maintaining the surfaces parallel to each other, so the rotor would be equally thick at all places. I started with 300-grain sandpaper, and during the process I worked down to 1200-grain one. At the end of the effort I had a rotor that was about 30 microns thinner than at the start, parallel to better than 2 microns, and mirror-smooth. I was satisfied.
Then I worked down the apex seals until they had the same dimension as the rotor. In the process I noticed that these seals are surface treated: The first two or three microns were hard to get off, then the work proceeded much easier. I don't feel bad for having destroyed that hard surface at one end of each seal: The important place where a hardened surface is needed is where the seal contacts the external shell, sliding along it at high speed, pressed against it by centrifugal forces.
Now I had a clearance of about 50 microns between the rotor and the walls. So I had to adjust the external chamber, despite it being very well made. And first I had to decide how much clearance to leave! A hard decision, since I had no idea about the differential thermal expansion of the rotor and the chamber! I decided to go close, and break-in the engine carefully, so any needed clearance would build itself up. My goal was 3 microns.
It wasn't easy to achieve good flatness here, as the chamber is asymmetrical, having many more fins on the combustion/exhaust side than on the admission/compression one. But I finally managed to get it flat to better than 3 microns, with a clearance of 2 to 5 microns for the rotor.
I assembled my engine, and now it did have real compression! Excellent! I mounted it on the test bench, fueled up, connected the glow plug, turned it over to prime it - and the beast almost bit my finger off! It was running like a dream!
If you want to adjust a Wankel engine in this way, be warned that you
need to work at a constant temperature, and let each piece stabilize before
measuring it. A few degrees temperature difference can eat up several microns!
The extremes go on. This engine is extremely thirsty, using about 3 times as much fuel as a similarly powered 4-stroke piston engine. But the carburetor setting is extremely uncritical: A full turn barely makes any difference! It runs extremely smoothly: Almost no vibration, very good idling (at about 2500 to 3000 rpm). Unfortunately it is an extremely dirty engine too: It spills fuel and oil out of all bearings and joints, and as it also runs extremely hot, the spilled oil immediately bakes into the metal surfaces! Be sure to use castor oil in the fuel, there is simply no synthetic oil that holds up under the extreme working temperatures of this little devil!
For starting, the glow plug must be extremely hot. Red-glow is not enough
here! It needs to glow bright orange, almost yellow, otherwise the engine
won't start. But then, it starts easily! Contrary to what some people have
reported, it can be hand-started without trouble. Probably those who reported
needing an electric starter got Wankels with bad compression... But on
the other hand, an electric starter keeps your fingers safe. This beast
really likes to eat fingers!
Like a 2-stroke engine, it has one power stroke for each crankshaft revolution; no separate valves, the rotor controls inlet and outlet ports.
And like a 4-stroke engine, it fully separates admission, compression, combustion and exhaust; the crankcase is not used for precharging.
The OS Wankel sounds more like a 2-cycle model engine than like a 4-cycle
one, but specially at idle speed the sound is dominated by the grinding
noise of the rotor gearing. One has to become accustomed to the sound,
but I love it!
The problem of properly sealing a Wankel engine's moving parts is a tough one. While piston engines use multiple springy steel rings for sealing their round pistons against the round cylinders, and get reasonably good results from this simple and age-old technology, the problem is much harder in the Wankel. There are long, flat edges to be sealed. The corners formed by the apex seals and the rotor-to-wall seals are particularly hard to get airtight. And bad sealing not only makes an engine loose power, and burn more fuel, but it also causes lots of additional pollution - something that is unacceptable in today's world.
Fortunately the technology has brought great advances in this field. In the 1950s it was still very common that the Wankel cars failed because of broken seals, but nowadays some companies specializing in sealing have demonstrated Wankel engines that run almost forever without such breakdowns. So today it all boils down to Wankel seals being more costly than piston rings, but they are approaching the same level of reliability and effectiveness.
Another matter is fuel efficiency. If you take a certain amount of fuel, and burn it, you get a certain amount of energy, which can be removed from the burned mass as heat, or as mechanical power through expansion. In a real engine, both things happen, and the ratio between the two defines the fuel efficiency. In the interest of good efficiency, as much energy as possible should be removed by expansion. So it is in the interest of fuel efficiency that the combustion chamber be as closely spherical as possible, thus minimizing the conductive surfaces for a certain amount of enclosed volume. This ideal is approached reasonably well by the piston engine with properly shaped pistons tops and cylinder heads, but the Wankel is very far from this! Its long, thin combustion chamber, having lots of surface, really sucks the heat out of the working gases, producing severe losses. The Wankel is inherently inferior to the piston engine in fuel efficiency matters, because of this fact.
The only way a Wankel could approach a piston engine in efficiency would be by using materials that either are thermally nonconductive (impossible for now), or by operating the entire engine at such a temperature that little heat is extracted from the gases. We do have metals now that can work hot enough, but we do not have proper lubricants. All present-day engines are temperature-limited by the available lubricants. And when higher-temperature lubricants become available, they will first benefit piston engines, since these expose no sliding surfaces to the gases at the moment of highest compression, while Wankel engines do. Wankels would need much better lubricants than piston engines, to improve efficiency by the path of running hotter.
For the time being, the Wankel is a novelty, that can be rightfully applied in places where its specific advantages outweigh its disadvantages, but the great mass of internal combustion engines will keep employing pistons. And I will keep looking for a nice plane that deserves being powered by my Wankel!